WO2016019857A1 - 服务派发系统及方法 - Google Patents

服务派发系统及方法 Download PDF

Info

Publication number
WO2016019857A1
WO2016019857A1 PCT/CN2015/086075 CN2015086075W WO2016019857A1 WO 2016019857 A1 WO2016019857 A1 WO 2016019857A1 CN 2015086075 W CN2015086075 W CN 2015086075W WO 2016019857 A1 WO2016019857 A1 WO 2016019857A1
Authority
WO
WIPO (PCT)
Prior art keywords
service
information
order
service provider
driver
Prior art date
Application number
PCT/CN2015/086075
Other languages
English (en)
French (fr)
Inventor
胡志琳
刘章勋
封朋成
崔玮
王维
张凌宇
刘滢
罗文�
胡涛
Original Assignee
北京嘀嘀无限科技发展有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from CN201410379713.3A external-priority patent/CN104167093B/zh
Priority claimed from CN201410397679.2A external-priority patent/CN104156443B/zh
Priority claimed from CN201410409108.6A external-priority patent/CN104183118B/zh
Priority claimed from CN201410413040.9A external-priority patent/CN104157133B/zh
Priority claimed from CN201410418423.5A external-priority patent/CN104156868A/zh
Priority claimed from CN201410421805.3A external-priority patent/CN104183123B/zh
Priority claimed from CN201410437102.XA external-priority patent/CN104156489B/zh
Priority claimed from CN201410705608.4A external-priority patent/CN104574947A/zh
Priority claimed from CN201510020526.0A external-priority patent/CN104537502A/zh
Priority claimed from CN201510163063.3A external-priority patent/CN104715426B/zh
Priority to SG11201700895YA priority Critical patent/SG11201700895YA/en
Priority to US15/501,824 priority patent/US20170228683A1/en
Priority to KR1020187037289A priority patent/KR20190000400A/ko
Application filed by 北京嘀嘀无限科技发展有限公司 filed Critical 北京嘀嘀无限科技发展有限公司
Priority to MYPI2017000173A priority patent/MY188692A/en
Priority to KR1020177003867A priority patent/KR20180006871A/ko
Priority to EP15829451.2A priority patent/EP3179420A4/en
Publication of WO2016019857A1 publication Critical patent/WO2016019857A1/zh
Priority to PH12017500192A priority patent/PH12017500192A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06311—Scheduling, planning or task assignment for a person or group
    • G06Q10/063112—Skill-based matching of a person or a group to a task
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0613—Third-party assisted
    • G06Q30/0619—Neutral agent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2465—Query processing support for facilitating data mining operations in structured databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/29—Geographical information databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q50/40—
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/02—Services making use of location information
    • H04W4/023—Services making use of location information using mutual or relative location information between multiple location based services [LBS] targets or of distance thresholds
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2216/00—Indexing scheme relating to additional aspects of information retrieval not explicitly covered by G06F16/00 and subgroups
    • G06F2216/03—Data mining

Definitions

  • the present invention relates to the field of services, and in particular to a service provider oriented service Distribute systems and methods.
  • Big data is a collection of data that cannot be acquired, managed, analyzed, and processed with conventional tools over an affordable time frame.
  • people With the development of communication technology and the terminalization and intelligentization of location services, people generate a large amount of personalized and personalized information in real time through smart devices such as mobile phones, tablet computers and notebook computers, for example, current location, service requirements, Current activities, historical locations, historical service content, past events.
  • This information because of its large size, large number, and complex structure, constitutes big data and cannot be handled by traditional manual methods.
  • a service dispatching system comprising: an information receiving module configured to receive service providing information from a service provider and service request information from a service requester; The module is configured to store service providing information and service request information received by the information receiving module; and a processing module configured to calculate the service providing information and the service request information stored by the storage module to obtain a feature result: If the feature result satisfies at least one criterion, Determining that the service request information is dispatched to a service provider; if the feature result does not satisfy at least one criterion, determining that the service request information is not to be distributed to the service provider.
  • the service request information includes two geographical locations.
  • the service dispatching system further comprises an output module configured to push the service request information to the service provider when it is determined to dispatch the service request information to the service provider.
  • the processing module in the service dispatch system includes a criteria storage unit configured to store at least one criterion.
  • the service providing information in the service dispatching system contains its current information.
  • the current information of the service provider includes its current location and motion information.
  • the location and motion information of the service provider includes its location information.
  • the location and motion information of the service provider includes its speed information.
  • the speed information of the service provider includes its speed direction.
  • the information receiving module in the service dispatching system is further configured to receive information from an information source.
  • At least one criterion in the service delivery system is selected from the group consisting of: a parameter indicating a service provider's response to the service request information; a parameter indicating the activity level of the service provider; the service provider The distance between the resident geographic location and the two geographic locations; the current location of the service provider and the angle between the vector formed by the two geographic locations and the current speed direction of the service provider.
  • a service dispatching method comprising the steps of: receiving, at an information receiving module, service providing information from a service provider and service request information from a service requester; a module that stores service providing information and service request information received by the information receiving module; The module calculates the stored service providing information and the service request information to obtain a feature result: if the feature result satisfies at least one criterion, determining to distribute the service request information to the service provider; otherwise, determining not to the service provider Distributing service request information; wherein the service request information includes at least two geographic locations.
  • the service dispatching method further comprises the step of, when the service request information is dispatched to the service provider, the service request information is pushed to the service provider by an output module.
  • the service providing information in the service dispatching method contains current information of the service provider.
  • the current information of the service provider includes its current location and motion information.
  • the location and motion information of the service provider includes its location information.
  • the location and motion information of the service provider includes its speed information.
  • the speed information of the service provider includes its speed direction.
  • the service dispatching method further comprises the step of receiving, at the information receiving module, information from an information source.
  • At least one of the criteria for the service dispatching method is selected from the group consisting of: a parameter indicating a service provider's response to the service request information; a parameter indicating a degree of activity of the service provider; The distance between the resident geographic location and the two geographic locations; the current location of the service provider and the angle between the vector formed by the two geographic locations and the current speed direction of the service provider.
  • the service dispatching method further comprises the step of assigning an integral to the service request information by the processing module.
  • the service dispatching method further comprises the step of issuing a credit for the service provider after the service provider executes the service request information.
  • Figure 1 Structure diagram of the service delivery system
  • FIG. 1 Structure diagram of the information receiving module
  • FIG. 1 Structure diagram of the processing module
  • Figure 4 Structure diagram of the criteria storage unit
  • Figure 5a and Figure 5b Flow chart of the service dispatching system
  • Figure 6 Flow chart for obtaining the driver's resident point
  • Figure 7 Flow chart of the driver's resident point decision
  • Figure 8 Schematic diagram of the distribution of driver candidate resident points
  • Figure 9 Flowchart of obtaining a busy period of a service provider
  • FIG. 10 Flowchart of preprocessing of address information
  • Figure 11 Flow chart for rewriting address information
  • Figure 12 Flow chart of abbreviated address information
  • Figure 13 Order auction flow chart
  • Figure 14 Flow chart of the integration operation
  • Figure 15 Flow chart of the decision to go to work
  • Figure 16 Flow chart for determining the order of the shift
  • Figure 17a Flow chart of inactive driver judgment
  • Figure 17b Flowchart of order dispatch combined with the judgment of the inactive driver
  • Figure 18 Determining if the driver needs to cross the obstacle flow chart
  • Figure 19 Schematic diagram for judging whether two line segments intersect
  • Figure 20 Schematic diagram for judging whether two line segments intersect
  • Figure 21 Schematic diagram for judging whether two line segments intersect
  • Figure 22 Flow chart for determining the order
  • Figure 23 Schematic diagram of driver movement direction and order direction
  • Figure 24 Step-by-step order determination and display flow chart
  • Figure 25 Flow chart of the method for processing an order
  • Figure 26 Flow chart of the method for processing an order
  • Figure 27 Flow chart for generating a grab probability vector
  • Figure 28 Flow chart of the target restaurant decision
  • Figure 29 Flow chart of the target courier decision.
  • modules in a system in accordance with embodiments of the present invention, any number of different modules may be utilized and executed on a client and/or server.
  • the modules are merely illustrative, and different aspects of the systems and methods may use different modules.
  • service refers to a specific task or transaction that an individual or entity performs or performs with other individuals or entities.
  • the tasks or affairs involved may be in some physical form, such as food, drinks, or some non-physical form, such as haircuts, vehicle loading, house cleaning, beauty, and clothing cleaning.
  • a "requester” or “service requester” is used to represent a person or entity requesting or ordering a service.
  • supplier is used to refer to a person or entity that can provide services to a “requester” or “service requester.” E.g, City citizens can order fresh fruit online from fruit retailers.
  • the system communicates with citizens and fruit retailers to obtain information about service requests and service offerings to arrange services.
  • the content of the service is exemplified by a taxi service or other transportation service
  • the service provider is embodied as a taxi driver, a taxi company, an individual owning the vehicle, or other providing similar vehicle service.
  • the individual or entity, and the service requester is an individual who seeks a taxi or vehicle service.
  • the content of the service is exemplified in the form of a meal delivery service, in which case the service provider is embodied as a person providing food, drinks, a store, a restaurant, etc., and the service requester is avatar-required to order, order Individuals, teams, etc. of drinks.
  • composition and structure of the system 100 of the embodiment of the present invention will be described below from the perspective of each module in conjunction with FIG.
  • the system includes an information receiving module 110, a storage module 120, a processing module 130, and an output module 140.
  • the connection between modules in the system can be wired or wireless. Any module may be local to the system or remotely connected to other modules over the network.
  • the correspondence between the module and the module may be one-to-one or one-to-many centralized.
  • one processing module 130 is connected to multiple information receiving modules at the same time, and may also be many-to-many, for example, multiple.
  • the information receiving module exchanges information with a plurality of storage modules.
  • the storage module 120 and the output module 140 do not have to be all present, but may be traded off as the application scenario changes. For example, for some service delivery systems that rely only on real-time information, the storage module 120 can be discarded without affecting the operation of the entire system. As another example, for some service delivery systems that exist in the background and that only implement information calculation and processing, the output module 140 can be discarded.
  • the above examples are only for explaining that the storage module 120 and the output module 140 are not necessary modules of the system, and those skilled in the relevant art can improve and change the configurations of the above two modules as needed, and these improvements and changes are The spirit and scope of the invention are not departed.
  • System 100 receives information from one or more service requesters (201-20m) while also receiving information from one or more service providers (301-30n).
  • the information from the service requester includes, but is not limited to, order information associated with it (211-21m).
  • system 100 can also receive information from information source 400.
  • the service requester (201-20m) generally refers to any service requester that makes a request for a specific service within a certain spatial scope and time period; the service provider (301-30n) generally refers to any specific space range and time period.
  • Service provider of the service may be an individual who needs vehicle service on the road, an individual who has a fruit demand, an individual who needs to dine, an individual or entity that needs to ship the item, and the like; the service provider (301-30n) may be Taxi drivers, car drivers, fruit retailers, food service providers, couriers or messengers.
  • Information source 400 generally refers to a source that can provide information to system 100.
  • the information source 400 is used to store information related to the service, such as geographic information, weather conditions, traffic information, legal and regulatory information, news events, living information, living guide information, evaluation of the service provider, and background information of the service provider. , background information of the service requester, etc.
  • the information source 400 may exist in the form of a single central server, or in the form of a plurality of servers connected through a network, or in the form of a large number of personal devices.
  • the devices can connect the cloud server with the user through a user-generated content, such as uploading text, sound, image, video, etc. to the cloud server.
  • a large number of personal devices together form an information source.
  • the information source 400 may be a municipal service system including map information and city service information, a traffic real-time broadcast system, a weather broadcast system, a news network, or the like, or some information including a large amount of historical order information. Database, etc.
  • the information source 400 can be a physical information source, such as a common speed measuring device, a sensing device, such as a driver's vehicle speedometer, a radar speedometer on a road, and a temperature and humidity sensor.
  • the information source 400 may also be a source for obtaining news, information, road real-time information, etc., such as a network information source, including but not limited to Usenet-based Internet newsgroups, servers on the Internet, and days.
  • the information source 400 may be a system that stores a plurality of catering service providers in a certain area, a municipal service system including map information and city service information, a traffic road condition system, a weather broadcast system, a news network, and the like. Or some servers that contain a lot of food and beverage history order information.
  • the above examples are not intended to limit the scope of the information sources herein, nor to the scope of the examples.
  • the present invention can be applied to any service, any device or network capable of providing information related to the corresponding service. Can be classified as a source of information.
  • the information receiving module 110 receives information from the service requester (201-20m), the service provider (301-30m), and the information source 400, and inputs the above information to the processing module.
  • the manner of information collection includes, but is not limited to, wired or wireless, obtaining or inquiring information from a service provider, a service requester, and obtaining information from the information source 400 in a manner of receiving subscriptions and pushes.
  • Information obtained or queried by the service provider and the service requester including but not limited to, information relating to the service provider and information relating to the service requester.
  • information about the service including but not limited to, the specific content of the service, such as taxi service, personal car, shared car, express delivery, delivery, delivery, parking, advertising, laundry, maintenance, entertainment, performance Waiting; the way the service is provided, such as the door service, delivery, the right to use the goods (including rental equipment or equipment), etc.; the payment method of the service, such as cash transactions, online payment, transfer and remittance, etc.; service tip or reward; service Time limits, such as real-time services, appointment services, etc.
  • Information obtained from the information source 400 in a manner of receiving subscriptions, pushes including but not limited to: basic information of the service provider and/or service requester, such as age, gender, nationality, address, ethnic group, religious belief, education Degree, work experience, marital status, emotional status, language ability, professional skills, political inclinations, hobbies, favorite music/television programs/movies/books, etc.; physiologically relevant information of service providers and/or service requesters, such as Height, weight, waist circumference, chest circumference, hip circumference, BMI index, vital capacity, visual acuity, color weakness/color blindness, past medical history, family history, medical history, etc.; historical information about service providers and/or service requesters such as driver's license information , traffic violation records, drunk driving records, criminal records, credit records, etc.; service provider historical service information, such as historical service Service time and location, service frequency, service content, received service evaluation, service result; service order information in the service requester history, such as the time and place of the historical service order, the frequency of receiving the service
  • the storage module 120 is configured to store a criterion for the processing module 130 to calculate and determine information received by the information receiving module 100, and output the criterion to the processing module 130.
  • the storage module 120 may include one or more types of storage capable of storing information accessible by the processing module 130, including but not limited to common storage devices such as solid state storage devices (solid state hard disks, solid state hybrid hard disks, etc.). Mechanical hard drives, USB flash drives, memory sticks, memory cards (such as CF, SD, etc.), other drivers (such as CD, DVD, HD DVD, Blu-ray, etc.), random access memory (RAM), and read-only memory (ROM).
  • solid state storage devices solid state hard disks, solid state hybrid hard disks, etc.
  • RAM is, but not limited to, decimal counting tube, counting tube, delay line memory, Williams tube, dynamic random access memory (DRAM), static random access memory (SRAM), thyristor random access memory (T-RAM), and zero.
  • ROM is but not limited to: bubble memory, magnetic button line memory, thin film memory, magnetic plate line memory, magnetic core memory, drum memory, optical disk drive, hard disk, tape, early NVRAM (non-NVRAM Volatile memory), phase change memory, magnetoresistive random memory, ferroelectric random access memory, nonvolatile SRAM, flash memory, electronic erasable rewritable read only memory, erasable programmable read only memory, Programmable read-only memory, shielded heap read memory, floating connection gate random access memory, nano random access memory, track memory, variable resistive memory, and programmable metallization cells.
  • the storage device mentioned above is a few examples, and the storage device that the network storage device can use is not limited thereto
  • the storage module 120 can also be used to cache information in the process of two-way interaction between the information receiving module 110 and the processing module 130, and store historical information related to the order, for other information used by the processing module 130, such as history. Traffic information, historical weather data, etc.
  • the processing module 130 receives the information from the information receiving module 110, performs arithmetic processing, and obtains one or more determination results.
  • the operations performed by the processing module can be logic-based operations, such as NAND operations, or numerical-based operations.
  • the processing module can include one or more processors, which can be any general purpose processor, such as a programmed programmable logic device (PLD), or an application specific integrated circuit (ASIC), or a microprocessor. It can be a system chip (SoC) or the like.
  • the output module 140 is configured to receive the result of the operation judgment from the processing module 130, and push the result to the service provider, or output the result to a third party, such as a server of the information service provider, for storage by the server.
  • the judgment result is disclosed periodically or irregularly, or the judgment result is obtained or inquired by the service provider.
  • the manner in which results or output results are pushed including but not limited to, various wired or wireless communications.
  • FIG. 2 shows a block diagram of an information receiving module 110 in accordance with an embodiment of the present invention.
  • the information receiving module 110 includes a service requester information receiving unit 111, a service provider information receiving unit 112, and an outside world information receiving unit 113.
  • the three unit modules, the information flows are bidirectionally connected to each other.
  • the service requester information receiving unit 111 receives information from the service requester 20x
  • the service provider information receiving unit 112 receives information from the service provider 30y
  • the outside world information receiving unit 113 receives information from the information source 400.
  • the information acquisition between the 113 and the information source 400 may be based on the same communication method or may be based on different types of communication methods. These communication methods can be wireless or wired. Among them, wireless communication methods include, but are not limited to, IEEE 802.11 series standards, IEEE 802.15 series standards (such as Bluetooth technology and Zigbee technology), first generation mobile communication technologies, and second generation mobile communication technologies (such as FDMA, TDMA, SDMA, CDMA).
  • Electromagnetic induction includes, but is not limited to, near field communication technology.
  • the wirelessly connected medium may be of other types, such as Z-wave technology, other paid civilian radio bands, and military radio bands.
  • the methods of wired communication include, but are not limited to, the use of metal cables, optical cables, or hybrid cables of metal and optics, such as coaxial cables, communication cables, flexible cables, spiral cables, non-metallic sheath cables, metal sheath cables, Multi-core cable, twisted-pair cable, ribbon cable, shielded cable, telecommunication cable, twin-strand cable, parallel twin-core cable, and twisted pair cable.
  • the information content from the service requester 20x acquired by the service requester information receiving unit 111, the information content from the service provider 30y acquired by the service provider information receiving unit 112, and the information content from the information source 400 acquired by the external information receiving unit 113 , has been involved in the above description, and will not be described here.
  • the information receiving module 110 is divided into three sub-modules for convenience of description.
  • the above-mentioned service requester information receiving unit 111, service provider information receiving unit 112, and external information receiving can be received by those skilled in the art.
  • Both or all of the units 113 are integrated on a single component.
  • the service requester information receiving unit 111 and the service provider information receiving unit 112 are integrated on one electronic component for collecting information generated by the service requester and the service provider in real time.
  • an external information receiving unit 113 is disposed to receive external information and historical service information. All such changes and modifications should be considered as the scope of protection sought by this application.
  • FIG. 3 shows a block diagram of a processing module 130 in accordance with an embodiment of the present invention.
  • the processing module 130 includes a processor 131, a controller 132, and a criterion storage unit 133.
  • the processor 131 interacts with the controller 132 in two directions, and the criterion storage unit 133 interacts with the controller 132 in two directions.
  • the processing module 130 interacts with the information receiving module 110 and the storage module 120 in two directions.
  • the processor 131 is configured to determine a specific order based on the criterion information stored by the criterion storage unit 133 and the information from the information receiving module 110 or the storage module 120, and output at least one determination result.
  • the controller 132 can be used to implement control of the processor 131 and the criteria storage unit 133 to ensure that the processor performs the determination function correctly.
  • the criterion storage unit 133 can be used to store the criterion information related to certain judgment conditions, and use the criterion information for the judgment process of the processor 131.
  • the criterion information may be from the local, the information receiving module 110, or may be from the information source 400.
  • the system will decide that the service order can be dispatched to the service provider; if the information is judged not to be satisfied At a certain time, the system will decide that the service order will not be dispatched to the service provider.
  • the criterion storage unit 133 will be further described below.
  • processors, controllers, and criteria storage units are only described in terms of their respective functions. In practical applications, the same hardware device, such as an application specific integrated circuit, or a microprocessor, may be used. , or any other device or device that can perform computational processing information and make logical decisions, configured to perform all functions of the processor, controller, and criteria storage unit, or configured to execute processor, controller, and criteria storage The function of any two of the elements without departing from the scope of the invention as sought.
  • FIG. 4 shows a block diagram of a criterion storage unit 133 in accordance with an embodiment of the present invention.
  • the criterion storage unit 133 includes a service provider service range criterion 1331, a service provider activity degree criterion 1332, a grab probability criterion 1333, a route obstacle criterion 1334, an order route criterion 1335, and Other criteria 1336.
  • the criterion corresponds to different judgment modes, which respectively correspond to service scope judgment, service provider activity degree judgment, grab probability judgment, route obstacle judgment, pass order judgment, and other judgments. Break mode. This determination process is done in controller 132.
  • FIG. 5a is a flow diagram of a service dispatch system.
  • the order information from the service requester 20x is received by the information receiving module 110; in step S110, the driver current information from the service provider 30y is received by the information receiving module 110; in step S120, by the processing
  • the module 130 calculates and analyzes the current order information and the current information of the service provider to obtain a feature result.
  • the processing module 130 determines, based on at least one criterion in the criterion storage unit 133, whether the feature result is The predetermined at least one criterion is satisfied. If the result is YES, the process proceeds to step S140, and the order is dispatched to the service provider; if the result is no, the process proceeds to step S150, and the order is not distributed to the service provider.
  • the judgment process can also be based on multiple criteria, and parallel or serial structural relationships can be adopted between different criteria.
  • serial structural relationship even if only one of the criteria is judged to be no, the flow ends and the order is not dispatched to the service provider.
  • parallel structural relationship as long as there is a criterion that the judgment result is yes, the flow ends and an order is dispatched to the service provider.
  • the judgment process may also be a comprehensive judgment scoring process based on a plurality of criteria. By assigning respective weights to different criteria, a comprehensive score is obtained, and the score is compared with a preset threshold to determine the judgment. result.
  • the judging process may also be an optimization selection process according to the whole system result. After determining that a plurality of service providers are candidate service providers, further selecting criteria, continuing the judging/scoring process, and further screening the follow-up service providers, Until the final selection of one or more service providers optimized.
  • the feature result is generated based on the service request information, information from the service requester and the service provider. After being generated, the feature result is used by the processing module to compare and judge with the criteria stored by the criteria storage unit.
  • the manner of judgment includes, but is not limited to: comparing one feature value in the feature result with a threshold value in the criterion; determining whether a feature value falls within a range of values in the criterion; and selecting an feature value according to the criterion Sorting or assigning scores in the ranking rule/rating rule and comparing them with a threshold of order or score; determining whether the feature result triggers the criterion An event; determining whether the feature result meets a certain pattern in the criterion, and the like.
  • Figure 5b is a flow chart of a service dispatch system incorporating external information.
  • the order information from the service requester 20x is received by the information receiving module 110; in step S115, the service provider current information from the service provider 30y is received by the information receiving module 110; in step S125, The information receiving module further receives the information provided by the external information source 400.
  • step S135 the processing module 130 performs calculation and analysis on the current order information, the current information of the service provider, and the external information to obtain a feature result; in step S145, The processing module 130 determines, based on at least one criterion in the criterion storage unit 133, whether the feature result satisfies the preset at least one criterion, and if the result is yes, proceeds to step S155 to distribute to the service provider. The order; if the result is no, the process proceeds to step S165, and the order is not distributed to the service provider.
  • the service delivery system can fully obtain the information of the service requester and match the information of the service provider.
  • the information of the service requester may include information such as the location and motion information of the service requester, the current physiological/health state, the current mental state, the preference for the service provider, the preference for the service form or the content, and the like.
  • the positioning and motion information may include, but is not limited to, a current location, a current motion state, a current motion direction, a current motion speed, a current activity state, and the like;
  • the current physiological/health state may include, but is not limited to, hunger, satiety, Disease, blood pressure, pulse, heart rate, body temperature, electrocardiogram, brain wave, respiratory rate, blood sugar content, blood oxygen content, etc.;
  • preferences for service providers including but not limited to, in personal information, experience, other abilities or specialties Expectations or preferences in beliefs or political tendencies;
  • preferences for service forms or content may include, but are not limited to, information on service payment methods, home delivery, door-to-door delivery, shipping speed, and speed of pick-up.
  • the service provider's information may include information such as the service provider's location and motion information, current physiological/health status, current mental state, operational information, preferences for the service requester, and the like.
  • the positioning and motion information may include, but is not limited to, a current position, a current motion state, a current motion direction, a current motion speed, a current activity state, and the like; /health status, which may include, but is not limited to, hunger, satiety, disease, blood pressure, pulse, heart rate, body temperature, electrocardiogram, brain wave, respiratory rate, blood sugar content, blood oxygen content, etc.; operational information, which may include, but is not limited to, Current operational status, currently provided service content, service time, service geographic scope, service resident point, service object, service mode; preferences for service requesters may include, but are not limited to, personal information to the service requester, Cultural/educational/vocational information, skills and other preferences.
  • the information of the service requester or the service provider utilized may be one item or multiple items.
  • the information of the service requester may include information such as location and motion information of the service requester, current physiological/health status, current mental state, preference for the service provider, preference for the service form or content, and the like.
  • the positioning and motion information may include, but is not limited to, a current location, a current motion state, a current motion direction, a current motion speed, a current activity state, and the like;
  • the current physiological/health state may include, but is not limited to, hunger, satiety, Disease, blood pressure, pulse, heart rate, body temperature, electrocardiogram, brain wave, respiratory rate, blood sugar content, blood oxygen content, etc.;
  • preferences for service providers including but not limited to, in personal information, experience, other abilities or specialties Expectations or preferences in beliefs or political tendencies;
  • preferences for service forms or content may include, but are not limited to, information on service payment methods, home delivery, door-to-door delivery, shipping speed, and speed of pick-up.
  • the collection of the above information can be accomplished by a variety of smart devices or dedicated measurement devices. For example, information related to location or motion can be collected by devices with positioning functions; information related to health and physiology can be obtained. Smart wearable devices or medical devices with multiple sensors to collect; for mental state, expectations of service providers, and expectations of service forms or content, can be entered by the service requester on the smart device in the form of text, voice, etc. And collecting.
  • a service requester has a preference for a service provider, service content, or service form when seeking a service.
  • the service content itself is not unique, but There is some sort of substitutability, but there are still some personal tendencies for service requesters for a class of homogenization services that differ only in form. Exploring these tendencies can help achieve more efficient, more targeted, and more personalized service delivery.
  • preferences may include, but are not limited to, a requirement or preference for the service provider (driver) in personal information, professional experience, ability or specialty, belief, or political orientation.
  • the personal information may include, but is not limited to, gender, age, marital status, emotional status, educational level, and the like.
  • Experiences may include, but are not limited to, the driving age of the driver, the type of driver's license, the ranking of the driver's license, the driving model, the traffic accident record, the traffic violation record, and the passenger evaluation.
  • Ability or specialty including but not limited to, hobbies, language skills, sports expertise, etc.
  • Beliefs and political tendencies can include, but are not limited to, drivers' religious beliefs, political tendencies, party information, and social groups they participate in.
  • there may also be preferences or tendencies in the content or form of the service including but not limited to, performance requirements or preferences for the mounted vehicle, and content for the transportation service.
  • the performance requirements or preferences of the mounted vehicle may further include, but are not limited to, requirements for the brand, model, maximum speed, acceleration time, fuel consumption, horsepower, maximum acceleration, displacement, and emission standards of the vehicle.
  • Requirements or preferences for the content of the transportation service may include, but are not limited to, requirements for the driver to pick up the driver, and requirements for the transportation service experience.
  • the passenger's preference for the service provider may include the service provider's experience and ability or feature information, and further preferably, the experience and ability or expertise may include the service provider's driving age and language ability, for example, Brazilian tourists who come to Berlin may want drivers to have basic listening and speaking skills in Portuguese.
  • preferences may include, but are not limited to, preferences for service providers (restaurants and/or food delivery personnel) and preferences for the restaurant itself.
  • Preference for service providers may include, but is not limited to, business experience for restaurants (including but not limited to, opening hours, business scope, food and beverage hygiene, food and beverage hygiene incidents, customer reviews) Preference, preference for meal delivery speed, preference for food delivery staff or industry evaluation.
  • For dining Body preferences may include, but are not limited to, requirements for catering varieties, preferences for dietary tastes (such as acid, sweet, bitter, salty, fresh, etc.), preferences for the nutritional value of meals (eg calorific value, including carbon water) Requirements or preferences for major nutrients such as compounds, fats, proteins, minerals, and vitamins, and food safety levels (such as genetically modified foods, organic foods, and green pollution-free foods).
  • preferences for dietary tastes such as acid, sweet, bitter, salty, fresh, etc.
  • preferences for the nutritional value of meals eg calorific value, including carbon water
  • Requirements or preferences for major nutrients such as compounds, fats, proteins, minerals, and vitamins
  • food safety levels such as genetically modified foods, organic foods, and green pollution-free foods.
  • preferences for the service provider and preferences for the service content include preferences for the service provider and preferences for the service content.
  • the service requester's preference for the service provider may include, but is not limited to, the service provider's employment status (such as employment time, employment experience, employment qualification), service provider's customer evaluation, and the like.
  • Preference for service content may include, but is not limited to, preferences for laundry methods (eg, hand washing, dry cleaning, machine washing, etc.), dewatering methods (eg, centrifugal dewatering, machine drying, sun drying, etc.).
  • Mining the service requester's preferences can be implemented using the service requester's historical service information.
  • the historical service information contains personal information, professional experience, ability or expertise, beliefs or political inclinations of the service provider at the time, and may also contain specific information on the form or content of the service. This information can be obtained and analyzed by the processing module.
  • such information may include, but is not limited to, the personal information, experience, capabilities or strengths, beliefs, or political tendencies of the service provider (driver).
  • the personal information may include, but is not limited to, gender, age, marital status, emotional status, educational level, and the like.
  • Experiences may include, but are not limited to, the driving age of the driver, the type of driver's license, the ranking of the driver's license, the driving model, the traffic accident record, the traffic violation record, and the passenger evaluation.
  • Ability or specialty including but not limited to, hobbies, language skills, sports expertise, etc.
  • Beliefs and political tendencies can include, but are not limited to, drivers' religious beliefs, political tendencies, party information, and social groups they participate in.
  • historical transportation services may also have service content or form information, including but not limited to, information on vehicle performance information and transportation service content.
  • the performance information of the mounted vehicle may further include, but is not limited to, the brand, model, and most of the vehicle. Information on high vehicle speed, acceleration time, fuel consumption, horsepower, maximum acceleration, displacement, and emission standards.
  • the information of the transportation service content may include, but is not limited to, the driver's pick-up information, the transportation service experience, and the like.
  • information in a historical meal delivery service may include, but is not limited to, information of a service provider (restaurant and/or food delivery person) and information about the meal itself.
  • the information of the service provider may include, but is not limited to, the business experience of the restaurant (including but not limited to, opening time, business scope, food hygiene status, food hygiene incident, customer evaluation), and delivery Meal delivery speed, food delivery staff experience or industry evaluation.
  • the information of the food itself may include, but is not limited to, the variety of food, food taste (such as acid, sweet, bitter, salty, fresh, etc.), the nutritional value of the meal (such as calorie value, including carbohydrates, fat, protein, mineral elements, Indicators of major nutrients such as vitamins), food safety levels, such as genetically modified foods, organic foods, and green pollution-free foods.
  • food taste such as acid, sweet, bitter, salty, fresh, etc.
  • the nutritional value of the meal such as calorie value, including carbohydrates, fat, protein, mineral elements, Indicators of major nutrients such as vitamins
  • food safety levels such as genetically modified foods, organic foods, and green pollution-free foods.
  • the information in the historical laundry service may include, but is not limited to, information of the service provider and information of the service content.
  • the service provider's information may include, but is not limited to, the service provider's employment status (employment time, experience, qualification), service provider's customer evaluation, and the like.
  • Service content information may include, but is not limited to, laundry methods (hand washing, dry cleaning, machine washing, etc.), dehydration methods (centrifugal dewatering, machine drying, sun drying, etc.).
  • the ordering record of the individual for a period of time (for example, one year, half year, and one month) can be captured, the total number of historical ordering meals can be calculated, and the catering variety, eating taste, and catering included in each ordering record information can be captured.
  • the price and the speed of the meal are counted, and according to this, the most likely taste of the individual, the food taste, the food price, and the meal delivery speed can be obtained.
  • the possible preferences of a service requester can be predicted by categorizing and analyzing the demographic information, background information, and the like of the service requester.
  • the current service requester's possible preferences may be predicted based on the mainstream or common preferences of the same type of population in the historical service order data.
  • the division of such people can be based on multiple Different parameters or attributes, including but not limited to, personal information (age, gender, nationality, hometown, current address), cultural/educational/occupational information (educational level, university, university, cultural circle, occupation, employment) Unit), physiological information (health, body shape, blood type, height, weight), beliefs and political tendencies (religious beliefs, political tendencies).
  • the possible preferences of the current service requester can be predicted based on the personal information.
  • the personal information may include gender, age, nationality, and hometown.
  • the current service requester's possible preferences may be predicted based on educational or professional information.
  • the cultural/educational/occupational information may include, education level, education, degree, high school for study/graduation, college/study/graduate, major, academic circle, occupation, industry, occupation level , employment units, working hours, etc.
  • the current service requester's possible preferences may be predicted based on the physiological information.
  • the physiological information may include health status, height, weight, body type, visual acuity, blood type, and the like.
  • the current preferences of the service requester can be predicted based on beliefs and political tendencies.
  • the belief or political inclination may include a religious camp, a religious sect, a religious time, an in-teacher position, a political party, and a supported political camp.
  • the above information can be combined to jointly judge or predict the possible preferences of a certain body.
  • personal information and cultural/educational/occupational information can be considered together to accurately predict preferences.
  • the service delivery system will preferentially push his/her meal service request to the Sichuan restaurant or the Cantonese restaurant in New York City.
  • the service provider's information may include the service provider's location and motion information, current physiological/health status, current mental state, operational information, and service requester. Information such as preferences.
  • the positioning and motion information may include, but is not limited to, a current location, a current motion state, a current motion direction, a current motion speed, a current activity state, and the like;
  • the current physiological/health state may include, but is not limited to, hunger, satiety, Disease, blood pressure, pulse, heart rate, body temperature, electrocardiogram, brain wave, respiratory rate, blood sugar content, blood oxygen content, etc.
  • operational information which may include, but is not limited to, current operating status, current service content, service time, service Geographical scope, service resident point, service object, service mode;
  • preferences for service requesters may include, but are not limited to, preferences of service requesters in terms of personal information, culture/education/career information, skill characteristics, and the like.
  • the information from the service provider includes real-time information of the service provider, and further preferably, the real-time information of the service provider, including the current location, the current motion state, the current motion speed, and the current operational status.
  • the information from the service provider includes the current physiological state of the service provider, and more preferably, the current physiological state of the service provider, including current heart rate, blood pressure, body temperature, electrocardiogram, brain wave, blood sugar content, blood Oxygen content, etc.
  • the collection of the above information can be accomplished by a variety of smart devices or dedicated measurement devices. For example, information related to location or motion can be collected by devices with positioning functions; information related to health and physiology can be obtained. Smart wearable devices or medical devices with multiple sensors are collected; for mental state, current operational status, and preferences for service requesters, the service provider can input text, voice, graphics, etc. on the smart device. Acquisition can also be obtained by analyzing historical data.
  • the current location, motion state and motion speed information of the service provider can be collected by the device with the positioning function.
  • the device includes, but is not limited to, a smartphone, a tablet, an in-vehicle positioning device, a navigator, and the like.
  • the device acquires current position, motion state and motion speed information by using a positioning technology selected from the group consisting of Global Positioning System (GPS) technology, Global Navigation Satellite System (GLONASS) technology, and Beidou navigation system technology.
  • GPS Global Positioning System
  • GLONASS Global Navigation Satellite System
  • Beidou navigation system technology Beidou navigation system technology.
  • Galileo positioning system (Galileo) technology Galileo positioning system
  • QAZZ quasi-zenith satellite system
  • base station positioning technology base station positioning technology
  • Wi-Fi positioning technology various positioning and speed measuring systems that are provided by vehicles.
  • the physiological state of the service provider can be collected by means of a device with built-in sensors.
  • the sensors built in the device include, but are not limited to, a photoelectric sensor, a blood pressure sensor, an electrocardiographic signal sensor, a body temperature sensor, a heat flux sensor, a pulse wave sensor, a bioelectric sensor, a three-dimensional motion sensor, and the like.
  • the device includes, but is not limited to, a smart wearable device, an electrocardiograph, an electroencephalogram measuring device.
  • the smart wearable device includes, but is not limited to, a smart watch, a smart bracelet, a smart helmet, smart clothing, smart glasses, a health patch.
  • the scope of service is an important concept. Without loss of generality, the scope of service usually refers to the geographical area in which the service provider is willing to perform the service content. Service providers in different industries may have different service scopes. For example, taxi drivers usually use the main urban area of a certain city as a service area; catering service providers who provide take-out service may only provide delivery in certain blocks. The meal service, correspondingly, the several blocks mentioned constitute the service scope of the catering service provider.
  • a resident point is a place where a person or entity frequently appears, or where the person or entity is located for a long time.
  • the resident point may be the driver's home address, noon dining place; for a food service provider or a domestic service provider, the resident point may refer to the business or the individual's business location.
  • the service provider's service scope and resident point can be obtained by mining historical service information. If the historical service information includes the location information of the service, for example, the start/end point of the transportation service, the starting point/destination of the food service, the address/business place/service location of the service personnel of the housekeeping service, and the like. Through the collection of location information in a large amount of historical service information, the service delivery system can obtain the service provider's service scope/resident point information.
  • Figure 6 is a flow diagram of obtaining a service provider resident point.
  • the service provider resident point information may be input from the service provider, such as the service provider in the account associated with the system, recording his/her home address, work address, and frequently staying place. Or can be served by a computing device The analysis of the historical position of the provider and the information of the resident point obtained by the mining.
  • mining the resident points may be based on historical locations and historical trajectories from the service provider.
  • mining the resident point can include the following steps:
  • Step S200 Obtain service provider trajectory information in a preset time period to form a service provider information set.
  • the time period can be one week, one month, one quarter, half year, one year, but it can be longer or shorter.
  • the specific time period is not limited here. The longer the general time, the more accurate the calculation, but the calculation amount and the calculation.
  • the storage space required is also larger;
  • Step S210 Calculate a resident point of the service provider in the time period according to the service provider information set.
  • a service provider information set is a collection of information about a plurality of parameters of a service provider, wherein the plurality of parameters include, but are not limited to, a service provider identity (ID), a time when the information is reported (time stamp), and a service provider Location (including longitude, latitude) and, optionally, dwell time at the location.
  • the collection of information may also contain other parameters, such as the activity content of the service provider at the location, and the like.
  • the time the service provider stays in a location does not need to be explicitly provided to the system, and the system can calculate the dwell time at that location based on multiple driver information records.
  • the resident point of the service provider is calculated, and multiple candidate resident points can be obtained first, and then the candidate resident points are analyzed to obtain the final candidate point information.
  • a clustering algorithm can be employed.
  • the clustering algorithm customizes a distance and a period of time, and automatically classifies the latitude and longitude of the distance and time into a resident density area of a service provider according to the latitude and longitude and the dwell time in the service provider information set.
  • the Dbscan algorithm can be used to calculate the service provider resident point.
  • the algorithm for calculating the service provider resident point is not limited to the Dbscan algorithm, and there may be other methods, such as partitioning method (Partitioning).
  • FIG. 7 is a flow chart for determining the driver's resident point by using the processing module.
  • a time period T is preset, and the track information of a certain driver is acquired in the time period T, and a driver information set is formed.
  • the driver information set includes n latitude and longitude coordinates, and each The latitude and longitude coordinates are used as the candidate resident points A1, A2, ..., An of the driver;
  • the smart device may not know the location of the driver and upload the location data at any time, but only when the driver turns on the location function of the smart device or turns off the location function.
  • the candidate resident point may be the first location and/or the last location uploaded by the driver's smart device each day. For example, if the driver leaves the home every day and turns on the location function of the smart device, when the home is turned off, the function is turned off, and the first location coincides with the last location, which is the home address. Without loss of generality, in the above case, there may be two candidate resident points recorded every day. Some drivers may turn off the positioning function for a period of time, and then turn on the positioning function, so that there may be four candidate resident points recorded in one day.
  • the number of records of the candidate resident points per day is when the driver turns the positioning function on and off.
  • the number of latitude and longitude of the upload is correct.
  • the resident point calculated according to this method is generally the driver's home address, because the driver usually turns on or off the positioning function when leaving home or arriving home;
  • the smart device may know the location of the driver and upload location data at intervals, which may be a preset time interval or a random time interval.
  • the selection of the candidate resident point is not necessarily the latitude and longitude uploaded when the driver turns on or off the positioning function, but automatically updates the positioning function at intervals. All the uploaded places are used as candidate permanent points, and then stay at the point where the candidate resident point is greater than a threshold, because the resident point thus calculated may be the place where the driver pauses at lunch or other places, candidate resident
  • the setting method of the point is not specifically limited herein.
  • step S310 the total distance from each candidate resident point to other candidate points is calculated, and the candidate permanent resident point with the longest total distance is deleted.
  • step S330 is performed, otherwise the steps are repeated.
  • step S330 if the distance between any two candidate resident points is greater than a distance threshold, then proceeds to step S350, the driver is considered to have no permanent point, otherwise step S340 is performed;
  • Steps S300 through S330 are calculation steps of the resident density region.
  • Steps S320 to S330 can be understood as a method of denoising, and for all the resident point candidate points, the total distance to other points is calculated.
  • the point with the largest total distance can be removed as a noise point, because this point is farthest from the possible center point, the loop is denoised according to this method, and the reserved point is closer to the driver's resident point, and the loop can pass
  • the method is judged as follows: that is, all remaining candidate points, the maximum distance between the two is not more than a defined threshold, and the threshold is generally 1 km to 5 km.
  • the threshold is generally 1 km to 5 km.
  • FIG. 8 is a schematic diagram of the denoising of the processing module determining the resident point. The higher the point above the graph is filtered, the six points 501, 502, 503, 504, 505, and 506 are successively removed as noise points.
  • step S340 the resident point is calculated using the remaining candidate resident points.
  • the method of calculating the resident point may be to solve the average value of the latitude and longitude values of the candidate resident point.
  • the method of solving the average value may be an arithmetic mean value or a geometric mean value.
  • the setting of the candidate resident point may be based on a stay time at the candidate resident point
  • the candidate resident point may be further filtered by setting a time threshold.
  • a time threshold may be set, such as 20 minutes, if the driver stays at a candidate resident point. Always less than 20 minutes, delete the waiting choose a resident point.
  • the candidate resident point whose staying time of the candidate resident point is less than a time threshold may be deleted first, and the remaining candidate candidate point is calculated by calculating the geometric mean value.
  • the time threshold for determining the resident point varies for different drivers and different vehicles, such as, but not limited to, 15 minutes, 30 minutes, or longer.
  • the specific resident point judgment needs to be considered in combination with the driver's personal, vehicle situation, time, location, road conditions and many other factors.
  • the point where the stay time is within 5 minutes may be the location where the driver is in a traffic jam and cannot be used as a data for calculating the resident point.
  • this information can be used in the dispatch judgment of the service order.
  • the service provider's resident point information can be combined with the location information of the service order for the processing module to make a judgment. If a service order contains a geographic location near a service provider's resident location, the order may be a more convenient order for the service provider, so the order can be dispatched to the service. provider. For more in-depth details, see the description of the judgment of the ride in the following.
  • the service provider capable of excavating the resident point information is not limited to the driver, any other service provider, for example.
  • Merchants who provide catering services, couriers who provide delivery services, messengers, individuals or merchants who provide laundry services, individuals or merchants who provide housekeeping services, etc. may have similar geographical locations or resident locations, and accordingly, Their resident information may be obtained by the service delivery system through a similar method or process.
  • the system can learn historical service information, and through the data mining of the service information, further obtain the service provider's business hours and busy time information.
  • Business hours also known as service hours, refer to the time when the service provider provides and executes the service. During this time period, the service provider can respond to the service requester's service request and execute the service content.
  • a busy period refers to a period of time during which the frequency or load of the service provider's execution service reaches a certain threshold during the service provider's business hours, and it is difficult to respond to a potential service request. For example, although a food and beverage provider is open from 7:00 to 23:00 on weekdays, it is only 7:30-9:00 every morning, 11:00-13:00 noon, 17:00-18:30 in the evening, and at night. It is busy from 20:00-22:00, and correspondingly, these four periods constitute the busy time of the restaurant merchant.
  • a service provider usually equipped with a dedicated service order management computer, server, to store historical service order data.
  • the service dispatching system can use information such as big data mining and pattern analysis to obtain information such as business hours and busy hours of a specific service provider based on historical service order data acquired from the information source 400.
  • historical service order information of a certain period of time may be collected, and the time period is divided into a plurality of sub-periods according to a certain time interval, and the number of service orders in each sub-period is counted, and different children are compared.
  • the number of service orders during the time period which can identify the business hours and busy hours.
  • step S400 the service dispatching system retrieves all the service information of a certain service provider within a predetermined time period T through the historical service information database, thereby forming a service information set;
  • this period T is divided into m sub-periods, denoted as t 1 , t 2 , ..., t m ;
  • step S420 counting and counting the service information in each sub-period is n 1 , n 2 , ..., n m ;
  • step S430 by comparing different n i (1 ⁇ i ⁇ m), the maximum value n k in n i is found, and the corresponding t k is found ;
  • step S440 tk is taken as a busy period of the service provider.
  • the predetermined time period T may vary depending on the service content and the overall operating state of the service industry, and may be one week, one month, one quarter or even one year.
  • the division of the sub-periods is also similar, and the length of the sub-period can be half an hour, one hour, two hours, three hours, twelve hours, one day, one week, and the like. Usually the length of the sub-period is shorter than the length of the entire time period.
  • the division of the sub-periods t 1 , t 2 , ..., t m is not necessarily uniform, and the lengths of different sub-periods may be the same or different, for example, for the 7:00-9:00 time period, it may be divided into 7 : 00-7:30, 7:30-7:45, 7:45-8:00, 8:00-8:15, 8:15-8:30, 8:30-9:00 six sub-periods.
  • the time T can be set to one day, and each sub-period can be set to half an hour, and the driver's business situation in the half hour period within one day is examined, and the busiest is calculated.
  • Time period another preferred case is to set a week as the length of T, and each sub-period can be set to one day to examine the daily business of the driver within one week to calculate the busiest day.
  • the time T can be set to one year, and each sub-period can be set to one quarter, and the laundry service situation in each quarter of one year is examined, and the most busy one is calculated.
  • Season another preferred case, set the year as the length of T, each sub-period can be set to one month, to examine the monthly business conditions within one year to calculate the busiest January.
  • the time period with the largest number of services is used as the busy period in this example, this standard is not unique.
  • the top three or the top five sub-periods of the order number can be used as three or five. Busy hours.
  • the services provided by service providers have certain characteristics in terms of form and content.
  • their service characteristics may include, but are not limited to, fast average speed, low air conditioning temperature in the car, punctuality, etc.
  • service features may include However, it is not limited to the variety of food and beverages provided, the taste of food and beverage (such as acid, sweet, bitter, salty, fresh and other tastes), food processing methods, food freshness, average meal delivery time, meal delivery methods
  • services for providing laundry services Providers preferences may include, but are not limited to, laundry methods (hand wash, dry cleaning, machine wash), dewatering methods (centrifugal dewatering, machine drying, sun drying), average laundry time, speed of delivery of clothing, and the like.
  • the characteristics of the service provided by a service provider can be mined, so that the service provider can be recommended and distributed with more targeted service request information.
  • the characteristics of the service provider's service content can be realized by using the service provider's historical service information. For example, according to the previous history of meal delivery information of a catering merchant, the main catering variety, food taste, catering price, meal delivery speed, etc. of the catering merchant can be excavated. Specifically, the meal record of the catering merchant for a period of time (for example, one year, half year, and one month) can be captured, the total number of historical meal delivery times is calculated, and the catering variety and diet included in the information of each order record are recorded. Statistics on taste, food price, meal delivery speed, and service evaluation can be used to derive the most likely catering variety, food taste, food and beverage price, meal delivery speed, and service evaluation of the restaurant merchant.
  • the user experience of the service provider when providing the service is different from the user experience when the service requester accepts the service.
  • the user experience of the service provider is mainly the number of service orders, whether the service order can be received or the number of service orders received.
  • Service providers with poor user experience are often difficult to receive service orders, so their revenue and service enthusiasm are also low. Therefore, judging the active status of service providers and performing targeted policy operations on inactive service providers, thus retaining these service providers with poor user experience, is necessary to improve the operational efficiency of the entire service delivery system. .
  • Obtaining the service provider's service activity can be achieved by using historical service information.
  • the service delivery system can receive information from the service provider, and can also receive historical information and statistical information about the service provider.
  • historical information and statistical information include historical online conditions, historical grabbing conditions, historical orders, historical service information, and the like.
  • the system can find out whether a service provider is an inactive service provider according to the online situation and/or the grabbing situation of each service provider.
  • a service provider can be a delivery person in a catering delivery service, and a transportation service. Taxi/private car driver, door-to-door service staff in the home beauty service, etc.
  • the service provider's online status and/or rush status can be obtained through a service application installed by the service provider, including but not limited to, a transportation service app installed on the personal smart terminal, installed on the computer. Ordering system, client for multiple service applications.
  • the service delivery system may The service provider is included in the list of inactive service providers.
  • the service provider's service record and the grab ticket can be stored on the server as part of the service provider's historical data.
  • machine learning and data mining methods using these massive historical data to train the model, you can apply the model to accurately estimate the probability of grabbing orders for each order that the service provider chooses to present.
  • the estimated service provider's grab probability is used for the distribution of service orders by the post-service dispatch system.
  • the service delivery system can receive current information and/or historical information and/or statistical information from a service provider of the information receiving module.
  • historical information and statistical information include historical online conditions, historical grabbing conditions, and historical order related information.
  • the system can determine the probability of grabbing orders for the order that the service provider chooses to present each time according to the online situation and/or the grabbing situation of each service provider.
  • a service provider may be a delivery service person in a catering delivery service, a delivery person in a city delivery service, a door-to-door service person in a home beauty service, and the like.
  • This method of processing an order can be accomplished by obtaining at least one characteristic of a historical order and a response of a service requester associated with the historical order (whether the order is grabbed, whether the order is successful); according to the history order associated with The service requester's response assigns a weight to the at least one feature; acquires a feature of the current order; and selects a current order in the current order that will be presented to the service requester based on the weight corresponding to the feature of the current order.
  • such features may include meal delivery distance, meal size, delivery fee, and the like.
  • the machine learning model can be a logistic regression model that is widely used in dichotomous problems; in some embodiments, it can also be a support vector machines model; in other implementations In the way, you can also choose to use other machine learning models based on the test results.
  • the current order refers to an order to be presented to or presented to the service requester.
  • the current order may be an order that has not been presented to the service requester, or an order that is being presented to some service requesters but not yet presented to other service requesters.
  • the current order can be obtained from an online server.
  • the manner in which an order is obtained may include receiving an order directly from a person who is placing a taxi to place an order or receiving an order forwarded by another intermediary (eg, a website server, etc.).
  • one or more service requesters associated with the current order may be selected among the plurality of service requesters as candidate service requesters to whom the current order will be presented.
  • a service requester within a certain range of the location at which the current order is sent may be selected as the candidate service requester.
  • the candidate service requester can also be selected based on other factors, such as the direction of travel of the service requester, and the like.
  • the selected candidate service requester can be further filtered.
  • the characteristics of the current order may also include: the distance between the location where the order is sent and the service requester or the distance between the location where the person waiting to take the taxi waits for the taxi and the service requester, the destination to be traveled in the order, the purpose of the order The type of land (for example, airport, hospital, or school), the road conditions around the destination of the order, or the number of times the order was presented.
  • the characteristics of the current order and service requester may also include: additional tips to be paid, time to wait, number of passengers, whether to carry bulky luggage, and the like. Further, as described with respect to historical orders, the characteristics of the current order may be determined directly from the content determined by the current order, or may be further indirectly determined by processing the determined content by the server. For the characteristics of the current order, the weight corresponding to the characteristics of the current order can be utilized, ie, the special order with the current order during the training phase The weight of the corresponding feature assignment
  • the distance parameter of the transportation service can be specified as the characteristic of the service order, analyze the response degree of a service provider to different distance transportation service orders, train the service distribution system, and estimate for a real-time service order.
  • the method includes: obtaining a distance between an origin and a destination of a current order; obtaining a probability of a grab for the historical order by the service requester, wherein a distance between the origin and the destination of the historical order and the current The origin of the order is related to the distance between the destinations; and based on the probability of the grab, the current order is sent to the service requester.
  • the service requester can include both traditionally driving a vehicle, a ship, an aircraft service provider, and a vehicle for carrying/loading when unmanned.
  • the location of the origin of the current order and the location of the service requester can be separately obtained, and then the distance between the origin of the current order and the location of the service requester can be calculated.
  • the origin of the current order may be obtained from the order information; the location of the service requester may be determined by positioning system positioning information and/or base station positioning information in the service requester device of the service requester.
  • the distance between the origin of the current order and the location of the service requester may be a straight line distance between them, or may be reference route information and road condition information when they are placed in the navigation system. The actual travel distance of the vehicle calculated from the road information.
  • the probability of the order of the service requester for the historical order wherein the distance between the origin of the historical order and the service requester and the distance between the origin of the current order and the service requester Related. For example, when there are a large number of historical orders such that the distance between the origin of each historical order and the service requester is related to the distance between the origin of the current order and the service requester, it can be separately obtained.
  • the service requester may also obtain the probability of robbing the order of the historical order for the historical order.
  • the correlation between the origin of the historical order and the service requester and the distance between the origin of the current order and the service requester may be as follows:
  • the distance between the origin of the historical order and the service requester is equal to the distance between the origin of the current order and the service requester.
  • the service requester for the history The order grab probability of an order is largely equal to the service requester's probability of grabbing the current order. That is to say, if the service requester has a low probability of grabbing the order for the historical order, the service requester's probability of grabbing the current order is also largely low. Thus, the current order will likely be of no value or low value to the service requester, so the transmission of the current order will likely affect the delivery of high value orders for the service requester. Therefore, in this embodiment, by reducing the transmission of the current order with a lower probability of the service requester being robbed, it is possible to ensure fast and accurate delivery of high value orders for the service requester.
  • the distance between the origin of the historical order and the service requester and the distance between the origin of the current order and the service requester belong to the same distance interval, wherein the distance interval is in accordance with each historical order
  • the origination is pre-allocated with the distance between the service requesters.
  • 0-100 meters is the first distance interval, represented by P1; 100-200 meters is the second distance interval, represented by P2; 200-300 meters is the third distance interval, represented by P3; and so on .
  • the service is because the distance between the origin of the historical order and the service requester and the distance between the origin of the current order and the service requester belong to the same distance interval.
  • the requestor's probability of grabbing the historical order is largely approximated by the service requester's probability of grabbing the current order. That is to say, if the service requester has a low probability of grabbing the order for the historical order, the service requester's probability of grabbing the current order is also largely low.
  • the current order will likely be of no value or low value to the service requester, so the delivery of the order will likely affect the delivery of high value orders for the service requester. So here In one embodiment, by reducing the transmission of the current order with a lower probability of the service requester being able to grab a single ticket, it is possible to ensure fast and accurate delivery of high value orders for the service requester.
  • the information input to the system needs to be processed in advance.
  • the format of the input information is very It is difficult to unify; or, because of the uncertainty or non-standardization of information input, the input information is difficult to recognize, such as handwriting that is difficult to recognize in handwriting, dialects, accents, and image elements that are difficult to recognize in image input. . Therefore, in the service dispatching system, in order to process the subsequent processing steps of the module for calculating, judging, matching, etc. of the input information, generally, the input information needs to be preprocessed.
  • the preprocessing method generally includes the identification and extraction of feature information, and the conversion of information formats.
  • Preprocessing of the input information may include pre-processing of the input text information.
  • the text information generally, includes address information input by the service requester, identity information, order requirement information, and the like.
  • the service delivery system After receiving the information input by the service requester, the service delivery system needs to identify the input information, extract the feature information, and convert the information format. For example, in many scenarios that rely on geographic location information, the service provider and the service requester have different propensity to express the same place, and the understanding is different, which makes it difficult for the service requester to express or indicate the location of the service provider. Understand that it is difficult for service requests to be accepted or executed by the service provider.
  • a service dispatching system capable of rewriting or converting an address information (also referred to as a point of interest, or poi) input by a service requester and/or a service provider into a service provider that performs service content (commonly known as a service provider)
  • a service provider that performs service content
  • a single point of interest that can be understood in order to shorten the location of the docking single person, clear non-recognized locations (such as unknown communities), and provide a short and easy-to-understand description of the location.
  • the processing module preprocesses the input address information.
  • the service delivery system can convert the information to include a display name that serves as a logo (eg, New York City Public The library main building Stephen A. Schwarzman building) and a detailed address (eg, 42nd Street / Fifth Avenue in New York City) that already exists in the cloud server's address entry library, as input to the address rewriting module .
  • a display name eg, New York City Public The library main building Stephen A. Schwarzman building
  • a detailed address eg, 42nd Street / Fifth Avenue in New York City
  • Preprocessing of the input information may also include pre-processing the input voice information.
  • the voice information in general, includes the voice input by the service requester through the smart device, and the voice content may include a demand for the service, the identity information of the service requester itself, and the like.
  • the processing module needs to identify the input voice information, extract the feature information, and convert the information format. For example, in a food distribution service scenario, a takeaway user (service requester) calls a restaurant takeaway delivery person (service provider) by voice to inform the restaurant of the delivery person to deliver the meal time, the meal place, and the dish information.
  • problems such as tone, accent, sentence break, stuttering, etc. of the voice information existence may cause the delivery person (service provider) to be inaudible, incomprehensible or misunderstood.
  • the processing module is capable of identifying the voice information of the takeaway user (service requester) and extracting the feature information.
  • the voice information may be separated and recognized by the voice recognition device, so as to output the input voice as the corresponding text information, and perform pre-processing of the text information on the output text information.
  • the process of converting voice information into textual information commonly referred to as speech-to-text (STT).
  • STT speech-to-text
  • the steps of STT can be roughly divided into speech preprocessing and audio text recognition.
  • the speech recognition device can preprocess the audio by using a finite Fourier transform, a wavelet Fourier transform, a discrete Fourier transform, a convolution, a wavelet analysis, a filtering, a noise reduction, and the like, which are well known to those skilled in the art.
  • the processing module takes audio information including but not limited to pattern recognition, pattern comparison, pronunciation dictionary lookup, decoding, to output text information.
  • the processing module can also combine the language model, the grammar rules, and the historical data of the speech recognition with the existing mode to organize the text information into text content conforming to the grammar rules.
  • the preprocessing of the input information by the processing module may also include inputting image information Pretreatment.
  • the image information in general, includes photographs, videos, two-dimensional codes, and the like that are taken or input by the service requester through the smart device.
  • the image information may include service demand information, service requester personal information, service requester's own address information, identity information, and the like.
  • the processing module needs to identify the input image information, extract the feature information, and convert the information format. For example, in a home laundry service scenario, a laundry user (service requester) informs the laundry picker (service provider) of the specific condition of the laundry by taking a photo of the laundry.
  • the image information input by the service requester or the service provider may have problems such as poor definition, chromatic aberration, distortion, jitter, blurring, etc., which may cause the other party to have difficulty obtaining useful information.
  • the processing module can identify the picture information of the laundry pick-up person and extract the feature information.
  • a typical service order contains multiple address information, such as the current location of the passenger, the current location of the driver, the destination of the passenger, and the like.
  • drivers and passengers have different propensity to express the same place and have different understandings.
  • drivers are familiar with business districts, road names, and landmark buildings, while passengers tend to have more detailed and accurate representations, such as community names, building names, business names, etc., which leads to the location of passengers.
  • the driver is difficult to understand, and is unwilling to grab the order where the location is, directly affecting the driver's willingness to grab the order.
  • the processing module is capable of rewriting or converting an address information (also referred to as a point of interest, ie, poi) input by a service requester (such as an individual who needs a transportation service) into Service providers (such as drivers) understand the preferred geographic points of interest in order to shorten the location of the service provider, clear non-recognized locations (such as unknown communities), and provide a short and easy-to-understand description of the location.
  • an address information also referred to as a point of interest, ie, poi
  • Service providers such as drivers
  • the processing module preprocesses the input address information.
  • the pre-processing process is used to convert the input single geographic interest point text into two texts of a display name and a detailed address, so as input information of the address rewriting module.
  • Figure 10 is a flow chart of pre-processing address information.
  • step S500 receiving information including a display name and/or a detailed address
  • step S510 it is determined whether the information is a single text mixed with the display name and the detailed address; if yes, go to step S520; if no, go to step S630 to display the inclusion Display information about the name and the detailed address;
  • step S520 the information is cut into a display name and a detailed address
  • step S530 information including the display name and the detailed address is displayed.
  • step S510 of the pre-processing module if the display name of the geographic interest point is the same as the detailed address, or the display name is not empty, and the detailed address is empty, the text belonging to the geographic interest point is the display name and the detailed A single piece of text mixed with addresses.
  • the step S520 of the pre-processing module is to cut the display name or the detailed address of the geographic interest point into two records according to the area word, including the display name and the detailed address;
  • the area word is small, one, two, three, four, five, six, seven, eight, nine, ten, east, south, west or north, abandon the cutting;
  • the processing module further rewrites the input poi containing the display name and the detailed address into a poi that the orderer understands the preference.
  • the processing module can also rewrite the address information contained in the service request information sent by the service requester.
  • Figure 11 is a flow chart for rewriting address information.
  • step S600 receiving information poi including a display name and a detailed address
  • step S605 it is determined whether the poi contains a landmark building. If yes, clear the address, go to step S620; if no, go to the next step S610;
  • step S610 it is determined whether the displayname (display name) in the poi includes a common place name keyword such as a station, a village, a bridge, a subway, an interchange, or an airport; if yes, the address is cleared, and the process proceeds to step S620; if not, the process proceeds to a step S615;
  • a common place name keyword such as a station, a village, a bridge, a subway, an interchange, or an airport
  • step S615 it is determined whether the displayname (display name) in the poi includes a keyword such as a road, a street or a road, or whether the length of the display name is greater than 8 Chinese characters; if yes, the processing is not processed, and the process proceeds to step S620; Then, go to the next step S630 (displayname contains words such as road, street or road, indicating that it is enough to indicate its own position, such as displayname: "Zhongguancun Street No. 11"; when displayname is too long, it indicates that it is enough to indicate Your own location, such as displayname: "medium Guancun e world c, opposite Sinosteel International));
  • step S620 after clearing the detailed address, proceeding to step S625;
  • step S630 the latitude and longitude are reversely analyzed, and there is no latitude and longitude for the poi (for example, the passenger manually inputs "Zhongguancun" without the latitude and longitude information), and the latitude and longitude (poi) corresponding to the poi is parsed from the address latitude and longitude set by the display name or address of the poi. If there is latitude and longitude, ignore this step), calculate the starting point end distance;
  • step S635 it is determined whether the latitude and longitude distance of the starting point poi and the ending point poi is greater than the limiting distance (for example, the limiting distance is 50,000 meters to 60000 meters), if yes, no processing is performed, and the process proceeds to step S625; if not, the process proceeds to the next step.
  • S640 This step is to prevent the latitude and longitude error prevention steps, the latitude and longitude information is accurate, you can also ignore this step);
  • step S645 the detailed address is set as a business circle, and proceeds to step S625 to output the display name and the detailed address;
  • step S650 it is judged whether the area and the road information are acquired. If yes, the address is set to the area + road information, for example, "Changjingli Middle Street, Chaoyang District", go to step S655; if not, go to step S625, and output the display name and Address;
  • step S655 the process proceeds to step S625;
  • step S625 a poi containing the display name and the detailed address is output.
  • the specific method for determining whether the poi contains the landmark building in the step S605 is:
  • each record in the landmark building information set includes: a number, an address information, a click number, and the landmark building information set is used to collect the number of clicks of the passenger for each address information;
  • Each record in the address latitude and longitude set includes: number, address, latitude and longitude.
  • Each record in the latitude and longitude set of the quotient includes: number, business circle name and/or area road information, latitude and longitude.
  • the processing module is used to further shorten the address in the poi, and remove prefixes and suffixes that do not hinder the orderer's understanding, such as a city name prefix, a street number suffix, and the like.
  • Figure 12 is a flow chart of abbreviated address information.
  • step S700 input a poi including a display name and a detailed address
  • step S710 it is determined whether the prefix of the detailed address in poi is a city, and if so, proceeds to the next step; if not, proceeds to step S730;
  • step S720 the city prefix is cut off, such as Beijing, Zhangzhou City, Fujian province, because the existing order allocation is usually limited to a single city, so the city information is redundant;
  • step S730 it is determined whether the end of the detailed address in poi is the house number, if yes, go to the next step; if not, then go to step S750;
  • step S740 the last door number is cut off (for example, "Zhongguancun Street No. 11", and after processing, it becomes “Zhongguancun Street”; for the orderer, when listening to the order, it is mainly determined by the approximate area of the destination point whether there is a connection. Single demand, detailed location can be learned later);
  • a poi containing the display name and the detailed address is output.
  • the poi needs to be rewritten and abbreviated.
  • the above is a description of the pre-processing and calculation of the information input by the processing module based on historical data and service requester/service provider before dispatching the order.
  • the service delivery system can use this information, combined with specific criteria, for different Service request information is distributed to different service providers to achieve efficient and highly targeted service delivery.
  • the existing service system has the following two problems: 1.
  • the current service provider obtains the order mode as the grabbing mode, the smart device network condition is better, and the faster grabbing action is preferred to obtain the order, but the order is first grabbed to the order.
  • the service provider may be a service provider that is far away from the service requester or has a poor service.
  • the quality service provider cannot receive the order because the speed of the ticket is slow, and the service requester cannot obtain the service of the better service provider.
  • the order may be robbed by some medical institutions or individuals who have fast rushing orders but have poor medical effects or evaluations.
  • Cheating tools such as access speed and hardware performance undermine the fairness of platform competition.
  • This embodiment can be regarded as a dispatch system for obtaining a quality service provider based on an auction mode.
  • the system consists of two steps, a sorting step and an auction step.
  • the sorting step is configured to filter and sort the service provider set of an order according to factors such as an order characteristic indicator and a service provider information;
  • the auction step is configured to perform an auction grab order on the order in a service provider set that is filtered and sorted by the determining module.
  • the order has been processed by the processing module according to its characteristic index and n services
  • the factors such as the information of the service providers D1, D2, ..., Dn are calculated, and the order of a service provider is determined: D1>D2...>Dn, and Dx is one of the service providers.
  • the processing module assigns an auction time threshold T to the associated order, and an auction time tx is assigned to each service provider Dx, which may vary from service provider Dx.
  • the steps of the auction grabbing include:
  • Step S800 The driver Dx initiates a grab order
  • Step S810 Acquire an order of the driver Dx in the entire sequence D1>D2...>Dn;
  • Step S820 calculating an auction time tx
  • Step S830 determining whether Tx is 0, when Tx is 0, skipping to step S840, notifying Dx that the order is successful, and notifying the remaining driver that the auction failed, the auction is over, and when Tx is not 0, executing S850;
  • Step S850 Notifying the driver Dx+1 ⁇ Dn that the ranking is lower than Dx, the reason for the failure and failure of the auction, and notifying the drivers D1 ⁇ Dx-1 ranked higher than Dx to continue to participate in the auction, and the auction time is Tx;
  • step S800 When a driver having a rank higher than Dx participates in the auction, the process jumps to step S800 and loops.
  • a Beijing passenger uses a taxi software to book a taxi from Xizhimen to Sanlitun during work.
  • the passenger information obtained is:
  • any one of the distance, the driver's service quality and the driver's level is used as the weight, and the other two are the sub-options to sort the drivers;
  • Weight is a relative concept and is for a certain indicator.
  • the weight of an indicator refers to the relative importance of the indicator in the overall evaluation.
  • the weighting factor of the distance is 1, the weight of the driver's service quality and the driver's level is 0; if the weight of the driver's service is 1, the weight of the distance and the driver's level is 0; similarly, if the driver The weight coefficient of the level is 1, and the weight coefficient of the machine service quality and the distance is 0.
  • the driver's service quality can be directly derived from the passenger's historical evaluation of the driver (the passenger can score the driver every time he finishes the car);
  • the driver level is assessed by the number of orders received by the driver; for example, the level of the driver's order is less than 100, the level is D, and the level of 100 is less than 200, the level is C, the number is greater than or equal to 200, and the level is less than 300. , the rank of greater than or equal to 300 is A;
  • the distance between the driver and the passenger is prioritized, the service quality is the second item, and the driver level is the third item, and the following table is obtained:
  • the distance between the driver and the passenger is prioritized, and the driver level For the second item, the quality of service is the third item, sorted, and the following table is obtained:
  • the driver service quality is prioritized, the distance between the driver and the passenger is the second item, and the driver level is the third item, and the following table is obtained:
  • Driver number Quality of service (out of 100) Distance between driver and passenger Driver level 4 80 2.5km C 8 80 3km A 6 80 3.5km B 5 70 3km A 7 70 3.3km C 10 65 2.5km C 1 60 1.3km B 2 50 700m B 3 40 2km D
  • the driver service quality is the priority
  • the driver level is the second item
  • the distance between the driver and the passenger is the third item
  • Driver number Quality of service (out of 100)
  • Driver level Distance between driver and passenger 8 80 A 3km 6 80 B 3.5km 4 80 C 2.5km 5 70 A 3km 7 70 C 3.3km 10 65 C 2.5km 1 60 B 1.3km 2 50 B 700m 3 40 D 2km 9 40 D 3km
  • the driver level is the priority item
  • one of the service quality and the driver level is taken as the second item
  • the remaining item is sorted as the third item.
  • the specific ordering manner is not specifically limited herein.
  • driver service level The better the driver's historical service is, the more attractive it is to passengers. The better the driver's credit is, the better the order can be sold.
  • the auction time is negatively correlated with the sorting; sorting D1>D2>...>Dn, when the driver Dn grabs the order, the auction time increases, the driver who ranks higher than Dn can grab the order in the remaining time when the driver D1 grabs the order, the auction The time is shortened and the driver waiting for the auction will be released after the auction ends. Participate in the next auction.
  • the order is issued to 10 drivers
  • the auction time range is [0,7], divided into 10 segments
  • the top1 driver grabs the order after the auction time is 0
  • the top2 grabs the order after the auction time is 0.7s
  • ... top10 grabs the order
  • the post-auction time is 6.3s.
  • the order allocation it is recorded which drivers have been pushed by an order, and in the scoring module, the order-driver's score is recorded.
  • an order is pushed to 10 drivers and ranked as driver 1, driver 2, ..., driver 10 according to the quality level.
  • the auction process is as follows:
  • driver 1 is the first to grab a single
  • the driver is already the best driver and the auction ends directly. Tell other drivers that the order has been closed, the order is obtained by driver 1, and the match is the highest, and the passenger receives the message that driver 1 successfully grabs the order.
  • driver 1 For driver 1 to driver 4, if they grab the order, they may win the grab, so they will continue to have the opportunity to grab the bill within a certain period of time.
  • the processing module can assign appropriate points to the service order based on the value judgment,
  • the inclusion of points in the order auction affects the success rate of the service provider's rush to stimulate the service provider to undertake or execute service orders that are not easily accepted in other situations.
  • the general principle is that if a service request is difficult to accept by most service providers, a higher score is assigned to the service request; if a service request is accepted by most service providers, a lower score is assigned.
  • the technical solution adopted in this embodiment is: a service order crediting system for facilitating order transactions based on the order value judgment, comprising two steps: an order point allocation step and an exception processing step.
  • the order integral allocation determines the order value of an order according to historical data and performs corresponding point allocation; abnormal processing is used to perform behavior according to the service provider and the service requester after the order is robbed by a service provider And the relevant coordinate trajectory is judged to determine the point distribution.
  • the historical data is from an order set, and the order set includes: a call time, a service provider, and a service requester distance.
  • the order point allocation includes the following steps:
  • Step 1 preset a time period T, in which a plurality of time points T 1 , T 2 , . . . , T n are set , and the time points divide the time period into a plurality of time intervals 0 to T 1 , T 1 to T 2 , ..., T n-1 to T n ;
  • Step 2 Assign a number of grab orders to each time interval
  • Step 3 In each interval period, according to different grab orders probability, the corresponding score value is given to different orders, and the grab order probability comes from a pre-calculated probability set;
  • Step 5 assign a number of grab orders to each time interval
  • Step 6 In each interval, according to different grab orders, give the corresponding score value of different orders.
  • the time interval is further a certain time interval within the life cycle of the order.
  • the point allocation module further allocates a negative integral to the high value order obtained by the order judgment module, and allocates a positive integral to the low value order obtained by the order judgment module.
  • the order set is from a set of calling passenger information
  • the set of calling passenger information includes at least: a billing time, an order driving distance, and passenger position information.
  • the exception processing includes the following steps:
  • Step 1 After the driver grabs the order successfully, it begins to collect the geographical location of the driver and passengers;
  • Step 2 When the driver passes the service application and determines that the passenger has been received, according to the current driver and passenger geographic location, it is judged whether the service provider receives the passenger. If the service provider is confirmed to receive the passenger, the driver and the passenger's geographical position track are continuously recorded. ;
  • Step 3 When the passenger passes the service application to confirm that the destination has arrived or has completed the payment, according to the current passenger's geographical location, it is judged whether the driver delivers the passenger to the destination; if the passenger coordinates are at the destination, the points are issued; if the passenger coordinates Not at the destination, the order is considered a cheating order, deducting the driver's points.
  • FIG. 14 is a flow chart of the integration operation.
  • the order information is acquired, and the order information includes the starting position information of the transportation service, the ending position information, the time of requesting the service, and further, may include a request to the driver (service provider) side, such as driving Types of transportation (including models, such as cars, SUVs, Jeep; power types, such as electric, internal combustion engine power; fuel consumption; horsepower; brands, etc.), can include tip information, and can also include the reasons for requesting transportation services. Such as going to the airport by plane, going to the hospital, etc.;
  • driving Types of transportation including models, such as cars, SUVs, Jeep; power types, such as electric, internal combustion engine power; fuel consumption; horsepower; brands, etc.
  • Step S910 calculating the order credit, and calculating the points corresponding to the order according to the order information.
  • the points are affected by various parameters, such as the respective positions of the start and end points of the order, the length of the distance between the start point and the end point, and/or Distance, type of transportation, number of tips, urgency or cause of transportation services;
  • Step S920 receiving driver information for acquiring and executing the content of the order service, confirming After the service content is executed, the driver's information, such as age, gender, criminal record, integrity record, and employment time, can be obtained through the system to retrieve the cloud records;
  • step S930 it is determined whether an abnormal event has occurred. If not, the process proceeds to S940, the order has been completed, and the calculated order credit is issued; if yes, the process proceeds to S950, the order is not completed, and the calculated order credit is not issued.
  • the abnormal event refers to the unfinished traffic service caused by various reasons, such as the driver maliciously tricking the passenger to confirm the completion of the order, the driver and the passenger cheating. These anomalous events can be determined by the vehicle position and passenger position provided by the onboard positioning device and the positioning module in the passenger smart device.
  • the embodiment illustrates the flow of the integration operation by taking the transportation service as an example, it can be understood that other services, such as a meal delivery service, a laundry service, and a delivery service, can follow the similar integral operation to allocate points for the service, and After the service provider executes and completes the content of the service request, it issues points and deducts the points when the service provider's anomaly is discovered.
  • the transportation service order is generally pushed to all the taxis, and the taxi drivers are robbed without considering the individualized individual taxi drivers. demand. For example, when the driver is on the shift and receiving the shift, the driver is more inclined to meet the time-to-work time and space near the home; and the existing transportation service order allocation method does not consider the departure time of an order. Whether the departure location is associated with the time and place of one or some drivers to go to work and to receive the shift, but the order at this time and place is used as an order for the ordinary time and place order, so that the driver often does not get enough when going home. Or you can't grab an order with your own way.
  • a ride often referred to as a hitchhiker. If the direction of the transportation service sought by the passenger is the same or close to the driver's current or future direction, the passenger's transportation service demand constitutes a smooth order. Resident point, in this article, should be understood as the place where the service provider often stays and arrives within a certain period of time, such as family living. The address, the place of work of the person or spouse, the place of frequent dining, the school where the child is attending, the frequent gas station, the frequent gymnasium/sports hall/swimming hall, etc.
  • the order identification function based on the driver's resident point information can be realized, and the driver can push the driver to the resident point or leave the resident point, such as going to work, receiving the class, going out, etc. Orders that are closer to the distance, improve the targeting of the system to distribute orders, improve the efficiency and accuracy of the driver's orders.
  • This embodiment can realize the order judgment function based on the driver address information.
  • the order is a driver's order.
  • information from the driver including the driver's current location, driver's address information. Further, it also includes the driver number, the location of the shift or the position of the shift, the time of the shift or the time of the shift. Further, the current location of the driver can be obtained by a positioning module located in the smart device in the driver's vehicle and sent to the system.
  • the transportation service information required by the passenger includes an order number, a starting point position, an ending position, and a departure time.
  • the system determines whether the order is a driver's order by analyzing the information from the driver and the traffic service information requested by the passenger.
  • the information from the driver including the driver's current location, the driver's resident point and other information.
  • the driver's current location can be obtained by a positioning module located in the smart device in the driver's vehicle and sent to the system; the information of the driver's resident point can be obtained by historical data of the driver's location and running trajectory.
  • Figure 15 is a flow chart of the processing module performing the decision to go to work.
  • the specific function of the processing module 130 is to determine whether the traffic service sought by the current transportation service requester constitutes a shuttle order for a specific driver.
  • the processing module 130 according to the information from the information collecting module 110, specifically, the demand information of the transportation service from the passenger and the providing information of the transportation service from the driver, combined with the driver historical position, historical trajectory or input or system stored by the system
  • the calculated resident point information is used to determine whether the passenger's transportation service demand belongs to the driver's shuttle order.
  • the transportation service order information includes: an order number, a starting point position (such as a starting point latitude and longitude), an ending position (such as an latitude and longitude of the ending point), and a departure time.
  • the above information is not the entire order information, other information related to the order, for example, the reason for seeking transportation services, the preferences or requirements of the transportation service provider, for example, the age, gender, driving age, transportation of the transportation service provider Types, performance, etc., may also be included.
  • the taxi information includes: a driver number, a current location (such as the current latitude and longitude), a work location (such as the latitude and longitude of the work), or a stop position (such as the latitude and longitude of the work), the time of the work or the time of the shift.
  • a driver number such as the current latitude and longitude
  • a work location such as the latitude and longitude of the work
  • a stop position such as the latitude and longitude of the work
  • the above information is not the whole of the order information.
  • Other information related to the order such as the preferences or requirements of the traffic service requester, for example, the requirements or restrictions on the baggage carried may also be included.
  • the taxi information includes: a driver number, a current location, a work location, and an off-duty time;
  • step S1000 taking any car order Q in the car order set, taking any taxi information C in the taxi information set;
  • step S1010 the order Q departure time T, the order distance D, the taxi departure time Tout in the taxi information C, the taxi departure time threshold TYout, the departure point of the order Q and the taxi departure position are Dout, and the rental is obtained.
  • step S1020 if the absolute value of T minus Tout is less than or equal to TYout(
  • ⁇ TYout), it is considered that order Q satisfies condition 1 and proceeds to S1030; if T minus the absolute value of Tout is greater than TYout (
  • ⁇ TYout), it is considered that the order Q does not satisfy the condition 1, and the process proceeds to S250, and it is determined that the order is not the driver's outgoing order;
  • the order is the driver's outbound order; otherwise, enter S1050 and determine that the order is not the driver's outbound order;
  • the working time threshold TYout takes [0.5, 1.5] hours, that is, 0.5 hours to 1.5 hours. If the value of TYout is small, the number of matching orders that are selected is small or not. If the value of TYout is large, the driver's time of work is delayed or delayed.
  • the departure distance threshold DYout takes [3000,5000] meters, that is, 3,000 meters to 5,000 meters. If the value is small, the number of matched orders that are selected is small or not. If the value is large, the driver may have more distances.
  • the multiplication threshold Kout of the shift distance is [2,30] times, that is, 2 times to 30 times.
  • the value of the departure distance threshold DYout is fixed, the value of Kout is small, and the number of matching orders is more selected. Kout takes A large value will result in fewer or no matching orders.
  • Figure 16 is another flow chart of the processing module performing the determination of the windmill. In this flow, an order is judged to determine whether it constitutes a stop-and-go.
  • the taxi information includes: a driver number, a current location, a receiving position, and a time of receiving;
  • step S1100 take any car order Q in the car order set, take any taxi information C in the taxi information set;
  • step S1110 the order Q departure time T, the order distance D, the taxi information time T in the taxi information C, the taxi departure time threshold TYin, the starting point of the order Q and the current position of the taxi are Din1, and the rental is obtained.
  • the distance between the receiving position of the vehicle and the current position is Din2
  • the distance between the ending point of the order Q and the receiving position of the taxi C is Din3
  • the order distance of the order Q is D
  • the second receiving distance threshold DYin2 the shift distance multiple threshold Kin
  • step S1120 if the absolute value of T minus Tin is less than or equal to TYin(
  • ⁇ TYin), then the order Q is considered to satisfy the condition 1, and the process proceeds to step S1130; otherwise, Go to step S1160, and determine that the order is not the driver's order for the delivery;
  • steps S1120, S1130, and S1140 may be reversed. At the same time, these steps are not all necessary, and one or two judgment steps may be deleted as needed, without affecting whether the order is in the subsequent step. It constitutes the judgment of the smooth schedule. For example, it may be judged whether or not the condition in S1120 is satisfied, and it is determined whether the order Q is the on-time smoothing of the taxi C, and the judgment of the conditions in S1130 and/or S1140 is no longer performed.
  • the closing time threshold TYin takes [0.5, 1.5] hours, that is, 0.5 hours to 1.5 hours. If the value of TYin is small, the number of matching orders that are selected is small or not. If the value of TYin is large, the driver's time for receiving the shift is advanced or The delay is too long.
  • the first shift distance threshold DYin1 can have different value ranges.
  • the DYin value is [1000, 5000] meters, that is, 1000 meters to 5000 meters. If the value of DYin1 is small, the number of matching orders that are selected is large, and the value of DYin1 is large, so the driver has more distance to travel.
  • the second shift distance threshold DYin2 can have different values. As an example, DYin2 takes [3000, 5000] meters, that is, 3000 meters to 5000 meters. If the value of DYin2 is small, the number of matching orders that are selected is small, and the value of DYin2 is large, so the driver has more distance to travel.
  • the closing distance multiplication threshold Kin is [2,10] times, that is, 2 times to 10 times. If the value of Kin is small, the number of matching orders that are selected is large. If the value of Kin is large, the number of matching orders selected is small or No.
  • departure time threshold the departure distance threshold, the departure distance multiple threshold, the work time threshold, the first work distance threshold, the second work distance threshold, and the work distance multiple threshold are merely exemplary. Rather than limiting the corresponding concepts in the present invention, it is to be understood that any suitable threshold range is contemplated by those skilled in the art, and such contemplated threshold ranges are not departing from the scope of the claimed invention. spirit.
  • the processing module After calculating and determining the order for the specific driver, the processing module outputs the judgment result of the order to the output module for pushing to the designated driver, or provides the judgment result to a third party for the driver Query or get.
  • the driver's working time is 6:00 am, and the driver's address is No. 10 Zhongguancun Street.
  • the order format for a car that is collected from passengers is as follows:
  • Order number Passenger mobile number Departure departure time Longitude and latitude information of the place of departure 140002 13300000001 10 Zhongguancun Street 2014/2/2 06:00 Xxxxxx 140012 13300000002 20 Zhongguancun Street 2014/2/2 18:00 Xxxxxx
  • each taxi uses the smart device of the taxi driver to report the latitude and longitude of the current taxi location to the server at regular intervals (such as 10 seconds).
  • the information format is as follows:
  • the server will match the order and the driver.
  • the order 140002 belongs to the driver 12345, and the corresponding push information is set to “winding” when the order is broadcast. Order", prompt the driver.
  • the driver's time is 6:00 pm, and the address of the driver's home is on the 10th Street.
  • the format of the real-time call order collected from the passengers is as follows:
  • each taxi uses the smart device of the taxi driver to report the latitude and longitude of the current taxi location to the server every 10 seconds.
  • the information format is as follows:
  • the server will match the order with the driver.
  • the order 14012 is a stop for the driver 12345; the corresponding push information is set to “shunting order” when the order is broadcast, prompting the driver.
  • the smart device held by the driver can obtain information including driver identification information, real-time location of the driver, etc., and send it out periodically or irregularly.
  • the eight sets of information collected over a period of time collected by the driver are as follows:
  • the driver's trajectory is sent to the remote server by a message as described above.
  • the server can calculate the approximate resident point of the driver according to the Dbscan clustering algorithm. For example, the following table indicates that the driver will return to the following latitude and longitude when he is around 8:00 every night:
  • the location may be the driver's home address or the spouse's place of work.
  • the algorithm used to calculate the driver's resident point is not limited to the Dbscan clustering algorithm. It can be understood that the algorithm for calculating the driver's resident point may also include other methods, such as a partitioning method.
  • a partitioning method such as K-means, K-medoids, CLARA (Clustering LARge Application), CLARANS (Clustering Large Application based upon RANdomized Search), FCM, etc.; Hierarchical methods, such as BIRCH (Balanced Iterative Reducing And Clustering using Hierarchies), positive Binary-positive method, Rough Clustering Of Sequential Data (RCOSD); density-based methods such as OPTICS (Ordering Points To Identify The Clustering Structure); network-based methods such as STING (STatistical INformation Grid), CLIQUE (Clustering In QUEst), Wave-Cluster; model-based methods such as Cobweb, CLASSIT, etc.
  • the track information reported by a certain driver is stored.
  • the information collection format collected from the driver for a period of time is as follows:
  • the driver's approximate resident point can be calculated according to the Dbscan clustering algorithm.
  • the latitude and longitude of the residence time greater than 25 min is retained.
  • the driver's place for lunch at noon is near the following latitude and longitude:
  • the user experience when the driver uses the taxi software is different from the passenger.
  • the driver's user experience is mainly the order quantity, whether the order can be grabbed or how many orders are received.
  • For drivers with poor experience they often miss orders online and even feel that there is no real income. Therefore, some drivers will be lost due to poor experience. Therefore, judging the driver's online activity status and carrying out targeted strategy operations for inactive drivers, thus retaining these previously poor drivers, is necessary to improve the capacity of the entire system.
  • One implementation manner of this embodiment is to find out the driver who is easy to lose by screening the operating driver, and use the high quality order to recall the easy to lose driver.
  • the technical solution adopted in this embodiment is to determine whether each driver is an inactive driver according to the online situation and/or the grabbing situation of each driver within a certain period of time.
  • the specific function of the processing module is for a specific driver, based on Determine whether the driver is an inactive driver during the online situation and/or the grabbing situation of the driver for a period of time.
  • the robbing situation is received from the information receiving module, the car order history information set, and the order information in the car order history information set includes: an order number, a departure place, a destination, a departure time, and an order Driver number
  • the online situation is from a taxi information set, and each taxi information in the taxi information set includes: a driver number, a reporting time, and a taxi location;
  • Each inactive driver information in the set of inactive drivers includes at least driver number information.
  • Figure 17a is a flow chart of the processing module determining driver activity.
  • the inactive driver determination module comprises the following steps, as shown in FIG. 17a:
  • step S1200 any driver A is set to traverse the call order history information set and/or the taxi information set to query whether there is a record of driver A;
  • step S1210 if driver A has a record in the most recent T1 time period in the taxi information set, and driver A has not recorded in the most recent T1 time period in the call order history information set, then driver A is an inactive driver and enters Step S1250, and record the driver A to the inactive driver set; otherwise, proceed to step S1220;
  • the T1 time period may be set to different values, for example, 5 days, 7 days, 10 days, 15 days, 30 days, 90 days, etc., and may be other time lengths. The longer the time period is set, the smaller the influence of random factors on driver A's activity level judgment, but the amount of information will increase accordingly.
  • step S1220 if driver A has not recorded in the most recent T2 time period in the taxi information set, and driver A has recorded in the most recent T3 time period in the taxi information set, then driver A is an inactive driver, and proceeds to step S1250. And the driver A is recorded to the inactive driver set, otherwise proceeds to step S1230;
  • the T2 time period is 1 to 5 days
  • the T3 time period is 15 to 45 days. It can be understood that T2 and T3 can also be set to other different values, for example, 5 days, 7 days, 10 days, 15 days, 30 days, 90 days, etc., and other lengths of time. Similar to T1, the longer the time period is set, the random factor judges the driver A's activity level. The smaller the impact, the greater the amount of information.
  • step S1230 it is determined whether the average order record of the driver in the most recent T4 time period is not more than 1/10-1/2 of the average number of orders received by all the drivers in the latest T4 time period. If yes, the process proceeds to step S1250 to determine the driver. A is an inactive driver; if not, then go to step S1240, driver A is an active driver, and the T4 time period is 1 to 5 days;
  • step S1260 driver A is recorded to the inactive driver set.
  • the order information of the car is from a real-time information set of the car order, and each order information in the real-time information set of the car order includes at least an order number, a departure place, a destination, and a departure time.
  • Figure 17b is an order dispatch flow diagram incorporating a non-active driver decision.
  • the operation strategy implementation module traverses each car order information in the real-time information set of the car order and performs the following steps:
  • Step S1204 Obtain a car order information and an online driver A information in the real-time information collection of the car order;
  • Step S1214 determining whether the call order information is a good quality order, if not, proceeding to step S1234, playing the call order information to the driver, returning to step S1204, taking off the online driver information; if yes, proceeding to step S1224 ;
  • Step S1224 it is determined whether the order is locked, if yes, go to step S1264; if not, play the call order information to the driver, go to step S1234;
  • Step S1264 determining whether the lock time expires, the lock time expires, playing the call order information to the driver, returning to step S1234, playing the call order information to the driver; if the lock time has not expired, returning to step S1204, taking off An online driver information;
  • Step S1234 the order information is played to the driver A, and jumps to S1244;
  • step S1244 it is determined whether the driver A is an inactive driver. If not, the process returns to step S1204 to take down the online driver information; if so, the order is locked and the lock time is set, and the process returns to step S1204 to delete the online driver information.
  • the quality car order is screened according to the departure place, destination and/or departure time information of each order in the real-time information set of the car order.
  • the step of screening out a quality car order is:
  • the order Judging directly from the order destination, if the order destination is to go to the airport or train station, the order is a quality order;
  • the order distance of the order to judge is a quality order
  • X is between 3-10 km, other values are also possible, For example, 20km, 40km, 50km, etc.;
  • the inactive driver is actively involved in the order allocation, and the individual capacity and overall capacity of the taxi are also improved.
  • driver A has recorded in the most recent T1 time period in the taxi information set, and driver A has not recorded in the most recent T1 time period in the call order history information set, then driver A is an inactive driver;
  • the driver A's information is recorded to the inactive driver set.
  • driver A If driver A has not recorded in the most recent T2 time period in the taxi information set, and driver A has recorded in the most recent T3 time period in the taxi information set, then driver A is Inactive driver
  • the driver A's information is recorded to the inactive driver set.
  • Driver A has been offline for the last 3 days and has been online for the last 30 days, indicating that driver A is not active on the line and is also a situation of inactive drivers.
  • the T2 time period is 1 to 5 days
  • driver A has D records in the most recent T4 time period in the car order set. If D is less than or equal to one tenth of DX. To one-half (1/10 to 1/2), driver A is an inactive driver;
  • the driver A's information is recorded to the inactive driver set.
  • the number of orders received by Driver A is equal to one tenth of the average number of orders received by drivers. This indicates that Driver A’s willingness to take orders is much lower than the average number of orders received by drivers.
  • Driver A also belongs to a situation of inactive drivers.
  • the starting point of a certain order A and some The current position of a driver A is distributed on both banks of the Yangtze River, the linear distance is only 0.5km, and the driver A actually receives the order A from his current position. Passengers need to travel a long distance, cross the bridge, and then return for a distance. The actual journey may reach 5km. Therefore, driver A does not actually belong to the driver around the starting point of order A, and should not push order A to driver A.
  • This embodiment can overcome the shortcomings of the method of acquiring the driver around an order by the straight line distance in the process of assigning or pushing the car order, and avoid pushing the order to the driver who needs to cross the river, cross the river, and cross the overpass.
  • the specific function of the processing module is to determine whether each online taxi needs to cross an obstacle for a certain car order.
  • the processing module comprises the following steps:
  • Step 1 Collect at least two points on the obstacle to form more than one obstacle line segment, and take one of the obstacle line segments as P 1 P 2 ;
  • Step 2 Take any car order Q in the car order collection, set the starting position of Q to P 3 , and set more than one taxi to meet the conditions of the current location of the taxi around the starting point of the car order Q, forming a rental Car information collection;
  • Step 3 Take any taxi C in the taxi information set, and set the current position of C to P 4 to form the driver order line segment P 3 P 4 ;
  • Step 4 judging whether P 1 P 2 and P 3 P 4 intersect, if yes, the taxi C needs to cross the obstacle for the car order Q, and proceeds to step 6; if not, the next step;
  • Step 5 Whether each obstacle line segment has been traversed, if not, return to step 4, take an obstacle line segment to determine whether it intersects with the driver order line segment P 3 P 4 ; if yes, the taxi C does not for the calling order Q Need to cross the obstacle, go to step 6;
  • Step 6 whether the taxi information set has been traversed, if not, return to step 3, take the next taxi in the taxi information set; if yes, go to the next step;
  • Step 7 whether the car order set has been traversed, if not, return to step 2, and take the next car order in the car order set; if yes, end.
  • step of determining whether P 1 P 2 and P 3 P 4 intersect is:
  • condition 1 is satisfied, condition 2 is not satisfied, but When the starting position P 3 of the car order Q is on the P 1 P 2 line segment, P 1 P 2 and P 3 P 4 are intersected; if the starting position P 3 of the car order Q is in the P 1 P 2 line segment On the extension line, P 1 P 2 and P 3 P 4 are disjoint.
  • the obstacle is a river, a lake, a wetland, or an overpass.
  • the above technical solution of the present invention is based on the judgment of whether the line segments intersect, and can detect whether the driver around the order needs to cross the obstacle, and achieve the following technical effects:
  • the processing module determines a flow chart of the obstacle.
  • the processing module determining process includes the following steps:
  • Step S1300 collecting at least two points on the obstacle to form one or more obstacle line segments, the starting point is P 1 , the ending point is P 2 , and one of the obstacle line segments is P 1 P 2 ;
  • Step S1310 the order taking cab cab set to any one of order Q, Q is set as the start position P 3, it is provided with one or more taxi cab current position in line with the conditions around the cab line starting position Q to form a rental Car information collection;
  • step S1320 any taxi C in the taxi information set is taken, and the current position of C is set to P 4 to form a driver order line segment P 3 P 4 ;
  • step S1330 it is judged whether P 1 P 2 and P 3 P 4 intersect, and if yes, the process proceeds to step S1340, and it is determined that the taxi C needs to cross the obstacle P 1 P 2 for the car order Q; if not, the process proceeds to step S1350. It is determined that the taxi C does not need to cross the obstacle P 1 P 2 for the car order Q.
  • a vector-based cross product method can be used to determine whether the two line segments intersect, or other methods can be used to determine whether the two line segments intersect.
  • Figure 19 is a schematic diagram for judging whether two line segments intersect.
  • Fig. 20 and Fig. 21 are schematic diagrams for judging whether two line segments intersect in two special cases.
  • Fig. 20 and Fig. 21 are special cases in which the condition (1) is satisfied and the condition (2) is not satisfied because with Coincidence with The cross product is 0.
  • the cross product is 0, it should be discussed separately.
  • P 3 is on the line segment P 1 P 2 , the two line segments intersect; when P 3 is on the extension line of the line segment P 1 P 2 , the two line segments do not intersect.
  • condition (1) and condition (2) are satisfied at the same time, P 1 P 2 and P 3 P 4 intersect.
  • condition (1) is satisfied
  • condition (2) is not satisfied, but At this time, if the starting position P 3 of the calling order Q is on the P 1 P 2 line segment, P 1 P 2 and P 3 P 4 are intersected at this time; if the starting position P 3 of the calling order Q is at P 1 P 2 On the extension of the line segment, P 1 P 2 and P 3 P 4 are disjoint at this time.
  • the Yangtze River is approximately a straight line in Wuhan.
  • a driver is at No. 1125 Zhongshan Road, Jiang'an District, Wuhan City, Hubei province, with longitude and latitude (114.31108, 30.604891).
  • a passenger orders at the intersection of Heping and Qinyuan Road, latitude and longitude (114.338137, 30.59485), seeing the straight line distance between the driver and the passenger. Within 2km, and because the driver's passenger contacted across the Yangtze River, the driver needed to bridge the passengers, the actual distance was nearly 8km.
  • Two points (114.377997, 30.666914) and (114.157229, 30.380211) above the Yangtze River are used as the P 1 P 2 line segment, and the driver (114.31108, 30.604891) is connected with the order (114.338137, 30.59485) as the OD line segment, according to the above-described vector-based cross product method.
  • the method of judging whether the two line segments intersect can accurately identify that the P 1 P 2 line segment intersects with the OD line segment, so that the order is judged to be transmitted to the driver across the river, and the order is isolated from the driver, that is, the order is not broadcast to the driver. .
  • the driver's direction of travel and whether the order is on the way is an important factor affecting the driver and passenger experience.
  • the driver may hear orders within a certain range of the surrounding area, but there is often a problem that the driver's driving direction and the order orientation are inconsistent. For example, the driver is currently heading east, but the order for the driver is in the west direction relative to the driver. If the driver picks up the order, he needs to turn around to pick up the passenger. In the actual urban road environment, it may take a long time (consider allowing the turnover, traffic lights, overhead, etc.), seriously affecting the driver and passenger experience.
  • the angle between the driving direction of the driver and the order direction is taken as a factor affecting the ordering, and the driver is given priority to send the order in front of the driver's driving direction (shunting).
  • the embodiment discloses a method for processing an order, comprising: processing a module: receiving data related to an order; obtaining a direction of movement of the service provider; determining a direction of movement of the service provider and a direction from the service provider position to the order position The angle between them is less than a predetermined angle; and the data related to the order is sent to the service provider.
  • the predetermined angle is 90 degrees.
  • the method further comprises: determining that the congestion level from the carrier location to the order location is lower than the predetermined congestion level before transmitting the data related to the order.
  • the method further comprises: determining that the actual driving distance from the carrier position to the order position is less than the predetermined driving distance before transmitting the data related to the order.
  • the method further includes: determining, before sending the data related to the order, an angle between a direction from the carrier position to the predetermined position and a direction from the carrier position to the order position is less than a second predetermined angle .
  • the second predetermined angle is 90 degrees.
  • the method further comprises: increasing the order of the order in the order list before sending the data related to the order.
  • a method of processing an order in a carrier including: transmitting a direction of movement of the carrier; and between a direction of movement of the carrier and a direction from the carrier position to the order position When the angle is less than the predetermined angle, the data related to the order is received.
  • the predetermined angle is 90 degrees.
  • a method of presenting an order in a carrier including: receiving data related to an order; determining an angle between a direction of movement of the carrier and a direction from the carrier position to the order position Less than a predetermined angle; and presenting data related to the order.
  • the predetermined angle is 90 degrees.
  • the method further includes: before presenting the data related to the order, The congestion level from the carrier location to the order location is below the predetermined congestion level.
  • the method further comprises: determining that the actual driving distance from the carrier position to the order position is less than the predetermined driving distance before presenting the data related to the order.
  • the method further comprises: before presenting the data related to the order, determining that an angle between a direction from the carrier position to the predetermined position and a direction from the carrier position to the order position is less than a second predetermined angle .
  • the second predetermined angle is 90 degrees.
  • the method further comprises: increasing the order of the order in the order list before presenting the data related to the order.
  • This embodiment has the advantage of effectively reducing the driver's driving range and passenger waiting time, thereby improving the user experience.
  • Figure 22 is a flow chart for determining the order.
  • step S1400 data related to the order is received at the server. Get the position of the start and end points of the order, and calculate the direction of the connection between the start point and the end point.
  • the present invention may be implemented in a client-server architecture or implemented on a single device.
  • the manner of obtaining an order may include: receiving an order directly from a service requester that issues an order or receiving an order forwarded by another intermediary (for example, a website, etc.) through the information receiving module.
  • Obtain order-related data when the order is obtained including but not limited to the location of the person waiting to take the taxi when waiting for the taxi (referred to as the order location), the destination to be traveled, the extra tip that is willing to pay, the willingness to Waiting time, number of passengers, whether to carry large luggage, etc.
  • the person to be taxied above may be a software user who uses the taxi software to call the taxi, or another person who calls the taxi on behalf of the software user.
  • the order location may be represented by latitude and longitude coordinates, or may be represented by other information that may be used to indicate the determined location, including but not limited to buses. Stations, subway stations, a junction, and a specific building.
  • an order location is represented by information other than global positioning system coordinates, it can be converted to global positioning by the recipient of the order (eg, a server) or by a third party (eg, other address interpretation agencies such as a professional website) System coordinates for subsequent operations.
  • step S1410 the direction of motion of the carrier (driver) is obtained.
  • the direction of motion of the carrier is provided by the carrier, such as by a global positioning system.
  • the direction of movement of the carrier may be obtained by the server by obtaining the direction of movement of the carrier directly from the carrier or by other intermediary means (eg, a global positioning system information provider).
  • the direction of motion of the carrier obtained as described above is generally not separate direction information, but may include other information such as the carrier position (eg, the position represented by the global positioning system coordinates), the carrier's speed of motion, and the like. . Therefore, when other information such as carrier position or speed of motion needs to be used in other steps, the information required can be obtained in the manner obtained directly from the carrier or obtained by other intermediary means as described above.
  • step S1420 the angle between the direction in which the server determines the direction of movement of the carrier and the direction from the carrier position to the order position (referred to simply as the order direction) is less than the predetermined angle.
  • step S1420 the order direction is first determined. Since the order position and the carrier position can be obtained in steps S1400 and S1410, respectively, as described above, the order direction can be calculated by obtaining the longitude and latitude coordinates of the order position and the carrier position expressed by the global positioning system coordinates.
  • Figure 23 is a schematic diagram of the direction of movement of the driver and the direction of the order.
  • a direction 600 labeled N (North) is taken as a reference.
  • the server determines whether the angle ⁇ is less than the predetermined angle.
  • ⁇ when ⁇ is 0 degree, it indicates that the order position is exactly in the direction of movement 610 of the driver (carrier), and the carrier is the most convenient when going to the order position without considering the actual road, so this order
  • the direction is the best order direction for the carrier.
  • ⁇ when ⁇ is between 0 and 90 degrees, it indicates that the order direction is substantially in the same direction as the carrier's direction of motion 610. At this time, the carrier is generally on the way to the order position regardless of the actual road.
  • ⁇ 90 degrees, it indicates that the order direction is perpendicular to the carrier's moving direction 610.
  • the carrier does not follow the route or detour when going to the order position without considering the actual road.
  • ⁇ is greater than 90 degrees, it indicates that the order direction is substantially opposite to the carrier's direction of motion 610, and the carrier generally detours when going to the order position without considering the actual road. Therefore, only when ⁇ is less than 90 degrees, the carrier can go smoothly or substantially smoothly when going to the order position, so the predetermined angle can be set to 90 degrees. It should be understood that in accordance with other embodiments of the present invention, the predetermined angle may be set to a smaller angle, such as 45 degrees or less, when the carrier has a higher demand for the approach.
  • the server when the server determines that the angle between the direction of movement 610 of the carrier and the direction of the order is not less than a predetermined angle, the server may end the processing of the order for the carrier.
  • the predetermined travel distance may be a specific distance (such as 4 kilometers), or may be a distance referenced by a straight line distance from the carrier position to the order position (such as 1.5 times the linear distance) ).
  • step S1430 the processing module calculates that the angle between the direction of the line connecting the start point and the end point of the order and the direction of movement of the driver is less than the angle of the angle. If yes, judge the order The order is S1450 for the driver; if not, the order is judged to be the non-shunger order S1440 of the driver.
  • FIG. 24 is a schematic diagram of the process module determination and display flow of the processing module.
  • step S1500 the processing module acquires order information, including the positions of the start point and the end point, and calculates the displacement direction of the order;
  • Step S1510 the processing module acquires a moving direction of the driver
  • Step S1520 the processing module calculates an angle between the direction of the order displacement and the direction of movement of the driver
  • Step S1530 the processing module determines whether the angle is less than the angle threshold, and if so, determines that the order is the driver's order S1550, and displays the order information S1560 for the driver; if not, determines that the order is the driver's non- Take the order S1540.
  • the number of drivers and passengers using taxi software is increasing. How to achieve fast and optimal matching of large-scale online orders and drivers at the same time is a very challenging problem for algorithms and architecture.
  • the best match means that after considering various factors such as order characteristics, driver characteristics, surrounding driver volume, time, road conditions, etc., each driver is presented (for example, by voice broadcast or screen display) the current suitable order, and each The order has been fully presented.
  • the number of order broadcasts that the driver can hear is certain. If the driver does not have the accuracy to broadcast the order, that is, the order is not broadcasted, the valuable order broadcast channel is wasted, so that the driver is willing to choose (ie, willing to grab the order) and the order is not fully broadcasted, but also interferes with the driver. .
  • the daily order presentation log and the driver grab log can be stored in the storage module during the long-term operation.
  • the model can be applied to accurately estimate the probability of grabbing orders for the order each time the driver chooses to present.
  • the estimated driver grab probability is used for order allocation.
  • a method of processing an order comprising: obtaining at least one characteristic of a historical order and a response of a service provider associated with the historical order; Responding to the service provider associated with the historical order, assigning a weight to the at least one feature; obtaining a feature of the current order; and selecting a current in the current order to be presented to the service provider based on the weight corresponding to the feature of the current order Order.
  • obtaining at least one characteristic of the historical order and the response of the service provider associated with the historical order comprises: obtaining a response of the service provider selecting a historical order.
  • assigning weights to the at least one feature according to a response of the service provider associated with the historical order comprises: utilizing a machine learning model, according to at least one characteristic of the historical order, and whether the service provider selects a historical order The response, assigning weights to at least one feature.
  • selecting a current order in the current order to be presented to the service provider according to the weight corresponding to the feature of the current order comprises: utilizing a machine learning model, according to characteristics of the current order and with the current order The corresponding weight of the feature determines the probability that the service provider will select the current order.
  • selecting the current order in the current order to be presented to the service provider according to the weight corresponding to the feature of the current order further comprises: selecting the highest probability of being selected by the service provider in the current order Current order as the current order that will be presented to the service provider.
  • the machine learning model comprises a logistic regression model or a support vector machine model.
  • the method further comprises: updating the weight corresponding to the feature of the current order with the characteristics of the current order to be presented to the service provider and the response of the service provider associated with the current order.
  • the feature comprises at least one of: a location of the sending order and a service provider, a destination of the order, a destination type of the order, a road condition around the destination of the order, and The number of times the order was presented.
  • the historical order, the current order, and the service provider are associated with the same geographic area.
  • Figure 25 is a flow chart of a method for processing a module to process an order
  • step S1600 or S1610 may be performed during or before the training phase of the machine learning process.
  • steps can be performed in an online server or in a big data platform server.
  • the log can be obtained in the online server and stored as a historical sample-service provider identification pairing as a general sample.
  • the log may be retrieved from the online server and stored as a historical sample in the form of a historical order-service provider identification pair in another server, such as a big data platform server.
  • Step S1600 acquiring at least one feature of the historical order and a service provider response associated with the historical order; at least one feature of the historical order may include: an order fee, an additional tip to be paid, a location of the sending order, and a service provider's Distance, number of passengers, or the distance between the location where the taxi is waiting for the taxi and the service provider, the destination to be reached in the order, the type of order destination (for example, airport, train station, hospital or school) ), the road conditions around the destination of the order, or the number of times the order was presented.
  • the characteristics of historical orders may also include: the time to wait, whether to carry large pieces of luggage, the cause of the order, etc.;
  • Step S1610 Obtain a response of the service provider whether to select a historical order, the response includes a rushing behavior and a result thereof, such as whether to grab the order, the success or failure of the order, and the time when the order is received from the order (representing the speed of the order) Order willing), etc.;
  • Step S1610 is not required in itself, and can also be deleted from the entire process.
  • Step S1620 is performed after steps S1600 and S1610, and weights are assigned to at least one feature according to a response of the service provider associated with the historical order;
  • a step S1630 may be performed, where the machine learning model is used to assign weights to at least one feature based on at least one characteristic of the historical order and whether the service provider selects a response to the historical order. This step is also not necessary and you can assign weights directly to features.
  • the machine learning model can be a Logistic Regression model that is widely used in two-category problems.
  • the machine learning model can be a support vector machine model. In other embodiments, other machine learning models can also be used based on the test results.
  • ⁇ is the weight assigned to the feature, or the model parameter.
  • the predictor variable X and the target variable Y including the information of the plurality of sets of historical order-service provider pairings in the training data are substituted into the formula, and the weight ⁇ assigned to the at least one feature can be determined.
  • ⁇ may be a vector corresponding to a vector of the predictor variable X.
  • each feature can have a weight corresponding to the feature.
  • the location feature may have a weight of 0.5
  • the road feature may have a weight of 0.3
  • the order presentation feature may have a weight of 0.1.
  • the weights assigned to the features may be stored in a data file after the weights are assigned to the features.
  • the data file can be sent to an online server so that the online server loads the data file during the application phase to obtain weights.
  • step S1630 the process proceeds to step S1640, and the feature of the current order is acquired
  • the current order refers to an order to be presented to or presented to the service provider.
  • the current order may be an order that has not been presented to a service provider, or an order that is being presented to some service providers but not yet presented to other service providers.
  • the current order can be obtained from an online server.
  • the manner in which the order is obtained may include receiving an order directly from the person who is going to take the taxi that issued the order or receiving an order forwarded by another intermediary (eg, a website, etc.).
  • one or more service providers associated with the current order may be selected among the plurality of service providers as candidate service providers to whom the current order will be presented.
  • a service provider within a certain range of locations where the current order is sent may be selected as a candidate service provider.
  • the candidate service provider can also be selected based on other factors, such as the direction of travel of the service provider, and the like. In addition, you can also choose The candidate service provider is selected for further filtering.
  • the feature of the current order may be a feature corresponding to the feature to which the weight is assigned during the training phase.
  • the characteristics of the current order may also include: the distance from the location where the order is sent to the service provider or the distance between the location where the person waiting to take the taxi waits for the taxi and the service provider, the destination to be traveled in the order, the order The type of destination (for example, airport, hospital, or school), the road conditions around the destination of the order, or the number of times the order was presented.
  • Current order and service provider characteristics can also include: additional tips to be paid, time to wait, number of passengers, and whether to carry bulky luggage.
  • the characteristics of the current order may be determined directly from the content determined by the current order, or may be further indirectly determined by processing the determined content by the server.
  • the weight corresponding to the characteristics of the current order may be utilized, i.e., the weight assigned to the feature corresponding to the feature of the current order during the training phase.
  • step S1650 the current order to be presented to the service provider in the current order may be selected according to the weight corresponding to the feature of the current order;
  • step S1660 determining the probability that the service provider selects the current order based on the characteristics of the current order and the weight corresponding to the characteristics of the current order, this step is also not necessary;
  • the information of the feature in the current order-service provider pair composed of the current order and the candidate service provider associated with the current order may be used as a predictor.
  • the predictor variable and the weight represented by the vector corresponding to the feature in the predictor are substituted into the machine learning model for application.
  • the probability that the service provider will select the current order can be determined.
  • the machine learning model used in the application phase can be the same machine learning model as in the training phase. Type, such as a logistic regression model.
  • the machine learning model can be a support vector machine model.
  • the online server may compose a list of orders for a plurality of current orders received by the server at different times for one service provider for sorting.
  • the probability that the service provider selects the current order is compared to a predetermined threshold, if the probability is above a predetermined threshold, the current order is added to the order list for sorting; if the probability is below a predetermined threshold, the order list is not added.
  • the current order that the service provider obviously does not want to select can be filtered out.
  • the predetermined threshold may be stored in a configuration file of a respective program in the online server.
  • the threshold may also be adjusted periodically or dynamically based on factors such as the response of the service provider, the overall match status of the order at a particular time, and the like, as needed.
  • the service provider associated with the current order may be screened directly by probability without selecting the candidate service provider associated with the current order as described above.
  • step S1670 is next performed.
  • the current order with the highest probability of being selected by the service provider in the current order may be selected as the current order to be presented to the service provider.
  • the online server may sort the current orders according to the probability that the service provider selects a plurality of current orders, and then select the current order with the highest probability of being selected by the service provider as the forward direction. The current order presented by the service provider.
  • the current order can be sent to the service provider's client for presentation to the service provider.
  • a service provider's client-installed device eg, a mobile device
  • a user interface eg, a touch-sensitive display, etc.
  • the screen is displayed in a way to present the current order.
  • the service provider can select the current order as a response via the user interface depending on whether it is interested in the current order.
  • the online server may send a plurality of current orders to the service provider along with a probability that the service provider selects the current order, and implement multiple current orders on the service provider's client-installed device. Sort of. As can be seen from the above description, an order with a higher probability of being selected by the service provider can be presented, thereby achieving an order in which the priority presentation service provider prefers to select.
  • step S1640 may be returned to obtain a new current order for selection; or return to step S1650 Or S1660, based on the partially updated features, reselect the order that will be presented to the service provider.
  • the same current order may be presented to the same service provider multiple times due to the higher probability of being selected during its validity period. In this regard, the number of times the same current order is repeatedly presented to the same service provider can be recorded as a feature of the current order.
  • the weight corresponding to the feature of the current order is updated with the characteristics of the current order to be presented to the service provider and the response of the service provider associated with the current order. After step S1680, it may return to step S1620 to assign weights to at least one feature based on the response of the service provider associated with the historical order.
  • the grab probability is estimated for the order length.
  • the method includes: obtaining a distance between an origin of the current order and a destination; obtaining a probability of the service provider for a historical order, wherein the distance between the origin and the destination of the historical order and the current The origin of the order is related to the distance between the destinations; and the current order is sent to the service provider based on the probability of the grab.
  • the service provider can include both the traditional driving vehicle, the ship, the driver of the aircraft, and the passenger/loading when unmanned. Through tools.
  • the location of the origin of the current order and the location of the service provider can be separately obtained, and then the distance between the origin of the current order and the location of the service provider can be calculated.
  • the origin of the current order may be obtained from the order information; the location of the service provider may be determined by positioning information in the smart device of the service provider.
  • the distance between the origin of the current order and the location of the service provider may be either a linear distance between them or a reference route information or road condition information when they are placed in the navigation system. The actual travel distance of the vehicle calculated from the road information.
  • obtaining a ticketing probability of the service provider for the historical order wherein the distance between the origin of the historical order and the service provider and the distance between the origin of the current order and the service provider Related.
  • the service provider can also obtain the probability of the order of the service provider for the overall order of each historical order.
  • the correlation between the origin of the historical order and the service provider and the distance between the origin of the current order and the service provider may be as follows: (1) the history The distance between the origin of the order and the service provider is equal to the distance between the origin of the current order and the service provider.
  • the service provider for the history is because the distance between the origin of the historical order and the service provider is equal to the distance between the origin of the current order and the service provider.
  • the order grab probability of an order is largely equal to the service provider's probability of grabbing the current order. That is to say, if the service provider has a low probability of grabbing the order for the historical order, the service provider's probability of grabbing the current order is also largely low.
  • the current order will likely be of no value or low value to the service provider, so the transmission of the current order will likely affect the delivery of high value orders for the service provider.
  • the distance between the origin of the order and the service provider and the distance between the origin of the current order and the service provider belong to the same distance interval, wherein the distance interval is in accordance with the origin of each historical order. Pre-allocated with the distance between service providers. For example, 0-100 meters is the first distance interval, represented by P1; 100-200 meters is the second distance interval, represented by P2; 200-300 meters is the third distance interval, represented by P3; and so on .
  • the service is because the distance between the origin of the historical order and the service provider and the distance between the origin of the current order and the service provider belong to the same distance interval.
  • the provider's probability of grabbing the historical order is largely similar to the service provider's probability of grabbing the current order. That is to say, if the service provider has a low probability of grabbing the order for the historical order, the service provider's probability of grabbing the current order is also largely low.
  • the current order will likely be of no value or low value to the service provider, so the transmission of the current order will likely affect the delivery of high value orders for the service provider. Therefore, in this embodiment, by reducing the transmission of the current order with a lower probability of the service provider's grab, it is possible to ensure fast and accurate delivery of high value orders for the service provider.
  • FIG. 26 is a flow chart of a method of processing an order.
  • step S1700 the distance between the origin of the current order and the destination is acquired.
  • step S1710 is executed to determine the distance between the origin of the current order and the service provider, and the distance is attributed to the distance interval to which the order distance interval belongs; then, step S1720 may be performed to obtain a driver for the distance interval of the order.
  • the probability of a single order for a historical order which can usually be solved from the historical order pushed to the driver and the driver's response record for the historical order.
  • step S1730 it is determined whether the grab probability is greater than a certain probability threshold. If not, the process proceeds to step S1740, and the order information is not pushed for the driver; if yes, the process proceeds to step S1750 to push the order information for the driver.
  • FIG. 27 is a flowchart of a process for generating a sneak probability vector according to the present embodiment, including the following steps S1800 to S1840.
  • Step S1800 For each service provider, process the service provider's order information, and obtain, for example, the broadcast order distance and the grab order information of the historical order received in the past month, wherein the broadcast order distance refers to the origin of the order and The distance between the location of the service provider;
  • Step S1810 the plurality of historical orders are attributed to the plurality of order sets according to the distance interval
  • step S1820 the probability of grabbing for different broadcast distances is calculated. For example, for a broadcast distance of 0-100 meters, a broadcast distance of 100-200 meters, and a broadcast distance of 200-300 meters, the probability of grabbing is calculated separately.
  • the probability of a single order for a broadcast ticket distance of 0-100 meters may be equal to the number of grabs of a historical order received within a range of 0-100 meters in the past month and within a month The percentage of times the received order is received from a historical order in the range 0-100 meters.
  • a grab single probability vector is generated.
  • the distance intervals are assumed to be (A, B, C, D, E, F), the A interval is 0-10 km, the B is 10-20 km, and the C is 20-30 km. , D is 30-40km, E is 40-50km, F is 50-60km, and the single probability vector can be (0.1, 0.15, 0.2, 0.4, 0.3, 0.2). This vector indicates that the driver is 30-50km away. The order is most interested;
  • step S1840 the grab probability vector is stored.
  • the existing order dispensing system often requires the orderer to first determine a certain restaurant and enter the restaurant's ordering interface to operate. This requires the orderer to have a very clear understanding of the restaurant, and the demand for catering is relatively clear. However, in actual situations, there are often situations in which the orderer is unfamiliar with the surrounding environment, the restaurant's business conditions, consumption levels, and business practices are not known.
  • the embodiment provides a system for selecting a target restaurant based on analyzing order order information and restaurant information, and sending a order order to the target restaurant.
  • the order information of the order includes, but is not limited to, a name of the orderer, a reservation order, a delivery address, a meal delivery time, a cuisine, an acceptable price range, a preference taste of the orderer, etc.; the information from the restaurant includes the meal Hall opening hours, restaurant address, restaurant contact number, type of restaurant operated by the restaurant, usual consumption interval, etc.
  • the cuisines submitted by the order include but not limited to Sichuan, Cantonese, Shandong, Cantonese, Chinese, Su, Huaiyang, Zhejiang, Hunan, Anhui, French, Italian, Vietnamese, Mexican, Japanese. Korean cuisine, etc.
  • the orderer can select a certain dish in the order order, or can only give the type that he wants to eat.
  • the system selects the target restaurant according to the order record of the orderer and the taste of the preference, and the target restaurant can be one or more.
  • the specific food can also be just a summary of the cuisine, and the system recommends specific food for the orderer according to the order, preference, etc. of the orderer. Combined with other order information, select the target restaurant and send the order.
  • Figure 28 is a flow chart of the target restaurant decision.
  • step S1900 a order order Q in the meal order set is ordered, and a restaurant information C in the restaurant information set is taken;
  • step S1910 the information module 110 acquires the meal delivery address D1 of the order order, the meal delivery time T1, the meal type y, the acceptance price range is X1, and the restaurant business hours [T2, T3], the restaurant address D2, in the restaurant information C, The restaurant provides a collection of dining types of Y, the restaurant consumption price range is X2, and the distance threshold D;
  • step S1920 the processing module 130 receives the information sent by the information module and performs further determination processing. First, it is judged whether the distance between the restaurant address D2 and the meal delivery address D1 is less than or equal to the distance threshold D (
  • step S1930 it is determined whether the catering type y desired by the orderer belongs to the catering type set Y (y ⁇ Y) that the restaurant can provide. If yes, the process proceeds to step S1940. If not, the process proceeds to step S1960, where it is determined that the restaurant is not Target restaurant
  • step S1940 it is determined whether there is an intersection between the accepted price range X1 of the order order and the restaurant consumption price range X2. If there is an intersection, the process proceeds to step S1950, and the restaurant is determined to be the target restaurant; if there is no intersection Then, proceeding to step S1960, it is determined that the restaurant is not the target restaurant.
  • steps S1920, S1930, and S1940 can be reversed, and these steps are not all necessary.
  • One or two judgment steps can be deleted as needed, without affecting the target of the order order in the subsequent steps.
  • the judgment of the restaurant For example, it may be judged whether or not the condition in S1920 is satisfied, and it is determined whether the restaurant C is the target restaurant of the order order Q, and the judgment of the conditions in S1930 and/or S1940 is no longer performed.
  • step S1950 it is determined that the restaurant is the target restaurant, and the system may further transmit the information of the order order to the target restaurant.
  • the orderer first enters his own order request information through handwriting input, voice input and keyboard input through his mobile phone client.
  • the orderer user A
  • meal delivery time 11:30
  • delivery address Beijing Room 501, 5th Floor
  • Hailong Electronic Mall Zhongguancun
  • Haidian District with reference price: 15-25 yuan
  • catering category fast food
  • preference slightly spicy
  • the ordering software server has a large amount of restaurant information, and intercepts two of the restaurant information, as shown below:
  • the server matches the order and the restaurant based on the current order information and the restaurant information. According to the judgment rule of the target restaurant, the restaurant 0012 belongs to the target restaurant for the order 12001, and pushes the order order to the restaurant.
  • the embodiment provides a system for selecting a target courier based on analyzing express order information and courier information, and sending a courier order to the target courier.
  • the express order information includes the order start position, the delivery time, the end position, and the customer waiting time threshold. Values, etc.
  • information from the courier includes the current location, the delivery distance range preferred by the courier, the estimated time from the courier to the starting point of the order, and the threshold for receiving the order.
  • Figure 29 is a flow chart of the target courier decision.
  • step S2000 taking any express order Q in the express order information set obtained by the information receiving module, and taking any courier information C in the courier information information set obtained by the information receiving module;
  • step S2010 the starting point position D1, the ending position D2, the delivery time T1, the customer waiting time threshold T, the current position D3 of the courier from the courier information C, and the delivery distance range preferred by the courier are obtained from the courier order Q. [D4, D5], the courier to the order starting position estimated time T2, the order distance threshold D;
  • step S2020 the processing module receives the information from the information module, and performs a determination process. First, it is determined whether the distance between the courier current position D3 and the starting position D1 of the express order (denoted as
  • step S2030 it is determined whether the pick-up distance (
  • the starting point position of the order may be the current location of the sender, or may be a geographical location distributed from a foreign country to a certain distribution center;
  • step S2040 it is determined whether the deviation between the estimated time T2 of the courier from the current position to the start position of the order and the delivery time T1 is less than or equal to the customer waiting time threshold T (
  • steps S2020, S2030, and S2040 can be reversed, and these steps are not all necessary.
  • One or two judgment steps can be deleted as needed, without affecting the target delivery of the order in the subsequent steps.
  • the judgment of the staff E.g, It can be judged whether the condition in S2020 is satisfied, and whether courier C is the target courier of express order Q, and the judgment of the condition in S2030 and/or S2040 is no longer performed.
  • the system may further send the information of the order order to the target restaurant.
  • the sender first enters the application interface through his own smart device, and enters his courier request information by handwriting input, voice input or keyboard input.
  • the sending time 09:30 am
  • the sending address Beijing Haidian District Zhongguancun Hailong Electronic Mall, 5th Floor, Room 501
  • sender name User A
  • Tel 123...45
  • Address No. 20, Xueyuan Road, Haidian District, Beijing
  • to generate an order with the above information on the server of the courier service to generate an order with the above information on the server of the courier service .
  • the courier service server has a large amount of courier information, intercepting the two courier information, as shown below:
  • the server matches the order and the courier based on the current order information and the courier information.
  • the courier 0012 belongs to the target courier for the order 12001, and pushes the order order to the courier.
  • example embodiments may be described as a process embodied as a flowchart, a flowchart, a data flow diagram, a structure diagram, or a block diagram. Although a flowchart may describe the operations as a sequential process, many of the operations can be performed in parallel, concurrently or concurrently. In addition, the order of operations can be rearranged. When the operation of the process is completed, the process can be terminated, but the process can also have additional steps not included in the figure. Processes may correspond to methods, functions, programs, subroutines, subroutines, and the like. When the process corresponds to a function, the termination of the process may correspond to a return of the function to the calling function or the main function.
  • the processor and the memory are functionally illustrated as being in the same respective block, those skilled in the art will appreciate that the various devices, units, or steps described above may be This can be achieved by means of a general-purpose computing device, which can be concentrated on a single computing device, such as a service dispatching device, or not in the same device.
  • the information receiving module, the storage module, and the processing module may actually include a plurality of specific devices, such as wireless transceivers, memories, processors, microprocessors, etc., that are not located within the same physical enclosure.
  • the modules in the system according to the present invention may be distributed on a network composed of a plurality of computing devices, which may be connected by wire, may be connected by wireless in one area, or may be in different areas. Connected through a distributed network. Alternatively, they can also be executed by a computing device
  • the serial code is implemented so that they can be stored in a storage device by a computing device, or they can be fabricated into individual integrated circuit modules, or a plurality of modules or steps can be implemented as a single integrated circuit module.
  • the invention is not limited to any specific combination of hardware and software.

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Databases & Information Systems (AREA)
  • Educational Administration (AREA)
  • Development Economics (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Game Theory and Decision Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • Fuzzy Systems (AREA)
  • Remote Sensing (AREA)
  • Mathematical Physics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Software Systems (AREA)
  • Computational Linguistics (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Information Transfer Between Computers (AREA)
  • Telephonic Communication Services (AREA)

Abstract

一种服务派发系统及方法，该系统包含：一个信息接收模块（110），被配置为接收来自一个服务提供者（301-30n）的服务提供信息与来自一个服务请求者（201-20m）的服务请求信息；一个存储模块（120），被配置为存储信息接收模块（110）接收到的服务提供信息与服务请求信息；一个处理模块（130），被配置为，将存储模块（120）所存储的服务提供信息与服务请求信息进行计算，以得到一个特征结果：如果该特征结果满足至少一个判据（1331-1336），确定向服务提供者（301-30n）派发该服务请求信息；如果该特征结果不满足至少一个判据（1331-1336），确定不向服务提供者（301-30n）派发该服务请求信息。其中，服务请求信息包含两个地理位置。

Description

服务派发系统及方法
相关申请
本申请要求以下中国专利申请的优先权：
2014年8月4日递交，申请号为CN201410379713.3的中国发明专利申请，标题为“基于司机住址信息的顺风车运营系统”；
2014年8月13日递交，申请号为CN201410397679.2的中国发明专利申请，标题为“地理兴趣点模糊改写系统”；
2014年8月19日递交，申请号为201410409108.6的中国发明专利申请，标题为“基于拍卖模式获得乘客最优接驾司机的派单系统”；
2014年8月20日递交，申请号为CN201410413040.9的中国发明专利申请，标题为“基于司机在线活跃情况的运力拉升系统”；
2014年8月22日递交，申请号为CN201410418423.5的中国发明专利申请，标题为“基于订单价值判断促进订单成交的出租车积分系统”；
2014年8月25日递交，申请号为CN201410421805.3的中国发明专利申请，标题为“基于判断线段是否相交的叫车订单播送系统”；
2014年8月29日递交，申请号为CN201410437102.X的中国发明专利申请，标题为“基于司机轨迹进行司机常驻点挖掘的方法”；
2014年11月27日递交，申请号为CN201410705608.4的中国发明专利申请，标题为“处理订单的方法和设备”；
2015年1月15日递交，申请号为CN201510020526.0的中国发明专利申请，标题为“处理订单的方法和设备”；
2015年4月8日递交，申请号为CN201510163063.3的中国发明专利申请，标题为“用于处理订单的方法及设备”。
上述申请的内容通过引用形式被包含于此。
技术领域
本发明涉及服务领域，特别地，涉及一种面向服务提供者的服务 派发系统及方法。
背景技术
当前，随着智能设备，特别是智能定位、智能导航、智能手机的普及，给人们的生活方式带来了极大的便利。同时新兴的大数据系统、云计算系统，也为生活方式带来了新的变化。
大数据，是无法在可承受的时间范围内用常规工具进行获取、管理和分析、处理的数据集合。伴随着通信技术的发展，以及定位服务的终端化、智能化，人们每天通过手机、平板电脑、笔记本电脑等智能设备，实时地产生大量私人化、个性化信息，例如，当前位置、服务需求、当前活动、历史位置、历史服务内容、过往的活动。这些信息，由于其维度大、数量多、结构复杂，构成了大数据，已经无法通过传统的人工方法而处理。
上述技术的发展与应用为服务业带来了新的变革，产生了新的服务内容。例如，随着城市的发展，借助移动智能设备的交通服务已经是人们的普遍需求。又例如，日常生活中，常常需要各种基于定位的食物、货物的配送服务。在一些场景中，以上服务的服务请求者与服务提供者常常出现请求与供应信息的不平衡、服务内容的不明确、服务提供者与服务请求者之间通信不畅等情况，服务提供者收集、获取服务需求信息的及时性、准确性等受到影响。因此，难以为服务提供者派发适合的服务需求信息。
发明内容
根据本发明的一个方面，提供了一种服务派发系统，该系统包含：一个信息接收模块，被配置为接收来自一个服务提供者的服务提供信息与来自一个服务请求者的服务请求信息；一个存储模块，被配置为存储信息接收模块接收到的服务提供信息与服务请求信息；一个处理模块，被配置为，将存储模块所存储的服务提供信息与服务请求信息进行计算，以得到一个特征结果：如果该特征结果满足至少一个判据， 确定向服务提供者派发所述服务请求信息；如果该特征结果不满足至少一个判据，确定不向服务提供者派发所述服务请求信息。其中，服务请求信息包含两个地理位置。
根据本发明的进一步方面，该服务派发系统，进一步包含一个输出模块，被配置为当确定向服务提供者派发所述服务请求信息时，将服务请求信息推送给所述服务提供者。
根据本发明的进一步方面，该服务派发系统中的处理模块包含一个判据存储单元，被配制为存储至少一个判据。
根据本发明的进一步方面，该服务派发系统中的服务提供信息包含其当前信息。
根据本发明的进一步方面，该服务派发系统中，服务提供者的当前信息包含其当前定位与运动信息。
根据本发明的进一步方面，该服务派发系统中，服务提供者的定位与运动信息包含其位置信息。
根据本发明的进一步方面，该服务派发系统中，服务提供者的定位与运动信息包含其速度信息。
根据本发明的进一步方面，该服务派发系统中，服务提供者的速度信息包含其速度方向。
根据本发明的进一步方面，该服务派发系统中的信息接收模块进一步被配置为接收来自一个信息源的信息。
根据本发明的进一步方面，该服务派发系统中的至少一个判据选自由下列组成的群组：指示服务提供者对服务请求信息响应的参数；指示服务提供者的活跃程度的参数；服务提供者常驻的地理位置与两个地理位置的距离；服务提供者的当前位置与两个地理位置所构成的矢量与服务提供者的当前速度方向的夹角。
根据本发明的另一个方面，提供了一种服务派发方法，包含如下步骤：在一个信息接收模块，接收来自一个服务提供者的服务提供信息与来自一个服务请求者的服务请求信息；通过一个存储模块，存储信息接收模块接收到的服务提供信息与服务请求信息；通过一个处理 模块，将所存储的服务提供信息与服务请求信息进行计算，以得到一个特征结果：如果该特征结果满足至少一个判据，确定向服务提供者派发服务请求信息；否则，确定不向服务提供者派发服务请求信息；其中，服务请求信息包含至少两个地理位置。
根据本发明的进一步方面，该服务派发方法，进一步包含：当确定向服务提供者派发所述服务请求信息时，通过一个输出模块将服务请求信息推送给所述服务提供者的步骤。
根据本发明的进一步方面，该服务派发方法中的服务提供信息包含服务提供者的当前信息。
根据本发明的进一步方面，该服务派发方法中，服务提供者的当前信息包含其当前定位与运动信息。
根据本发明的进一步方面，该服务派发方法中，服务提供者的定位与运动信息包含其位置信息。
根据本发明的进一步方面，该服务派发方法中，服务提供者的定位与运动信息包含其速度信息。
根据本发明的进一步方面，该服务派发方法中，服务提供者的速度信息包含其速度方向。
根据本发明的进一步方面，该服务派发方法进一步包含在所述信息接收模块，接收来自一个信息源的信息的步骤。
根据本发明的另进一步方面，该服务派发方法中的至少一个判据选自由下列组成的群组：指示服务提供者对服务请求信息响应的参数；指示服务提供者的活跃程度的参数；服务提供者常驻的地理位置与两个地理位置的距离；服务提供者的当前位置与两个地理位置所构成的矢量与服务提供者的当前速度方向的夹角。
根据本发明的另进一步方面，该服务派发方法进一步包含：通过所述处理模块，为对服务请求信息分配积分的步骤。
根据本发明的另进一步方面，该服务派发方法进一步包含：当服务提供者执行服务请求信息后，为服务提供者发放积分的步骤。
附图说明
在此所述的附图用来提供对本发明的进一步理解，构成本发明的一部分，本发明的示意性实施例及其说明用于解释本发明，并不构成对本发明的限定。
图1：服务派发系统结构图；
图2：信息接收模块结构图；
图3：处理模块结构图；
图4：判据存储单元结构图；
图5a与图5b：服务派发系统流程图；
图6：获得司机常驻点的流程图；
图7：司机常驻点判定流程图；
图8：司机候选常驻点的分布示意图；
图9：获得服务提供者忙碌时段的流程图；
图10：地址信息的预处理流程图；
图11：改写地址信息的流程图；
图12：缩写地址信息的流程图；
图13：订单拍卖流程图；
图14：积分操作流程图；
图15：出班顺风车判定流程图；
图16：收班顺风车判定流程图；
图17a：非活跃司机判断流程图；
图17b：结合了非活跃司机判断的订单派发流程图；
图18：判定司机接驾是否需要跨越障碍物流程图；
图19：判断两线段是否相交的示意图；
图20：判断两线段是否相交的示意图；
图21：判断两线段是否相交的示意图；
图22：判定顺路订单流程图；
图23：司机运动方向与订单方向示意图；
图24：顺路订单判定、显示流程图；
图25：处理订单的方法流程图；
图26：处理订单的方法流程图；
图27：生成抢单概率向量流程图；
图28：目标餐馆判定流程图；
图29：目标快递员判定流程图。
具体实施方式
为了更清楚地说明本发明的技术方案，下文将对实施例描述中所需要使用的附图作简单的介绍。显而易见地，下文及附图所描述的仅仅是本发明的一些具体实施例，对于相关领域的技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些描述与附图，将本发明应用于其他方面。应当理解，给出这些示例性实施例仅仅是为了使相关领域的技术人员能够更好地理解进而实现本发明，而并非以任何方式限制本发明的范围。
虽然本文对根据本发明的实施例的系统中的某些模块做出了各种引用，然而，任何数量的不同模块可以被使用并运行在客户端和/或服务器上。所述模块仅是说明性的，并且所述系统和方法的不同方面可以使用不同模块。
本文中使用了流程图用来说明根据本发明的实施例的系统所执行的操作。应当理解的是，前面或下面操作不一定按照顺序来精确地执行。相反，可以按照倒序或同时处理各种步骤。同时，也可以将其他操作添加到这些过程中，或从这些过程移除某一步或数步操作。
如本文中所描述的，“服务”、“订单”、“服务订单”是指一个个体或者实体向其他个体或实体所履行或者执行的某种具体任务或者事务。所涉及的任务或者事务，可以是某种实物形式，例如，食品、饮品；也可以是某种非实物形式，例如，理发、车辆搭载、清扫房屋、美容、衣物清洁。“请求者”或“服务请求者”用以代表请求或订购某项服务的个人或实体。类似地，“供应商”、“提供者”、“服务提供者”用以代表可为“请求者”或“服务请求者”提供服务的个人或实体。例如， 城市市民可以在线向水果零售商户订购新鲜水果，此时，系统同时与市民以及水果零售商户进行通信，以获知与服务请求以及服务提供有关的信息，以安排服务。尽管在某些实施例中，服务的内容以出租车服务或者其他交通服务予以示例，此时，服务提供者具体化为出租车司机、出租车公司、拥有车辆的个人，或者其它提供类似车辆服务的个人或实体，而服务请求者化身为寻求出租车或者车辆服务的个人。在另外一些实施例中，服务的内容以送餐服务的形式予以示例，此时，服务提供者具体化为提供食物、饮品的个人、商店、餐厅等，而服务请求者化身为需要订餐、订购饮品的个人、团队等。但是，可以理解，这些示例并非出于限制本发明中服务、服务请求者、服务提供者等含义的目的，事实上，任何涉及到一个个人或实体与另一个人或实体之间的有形或无形服务，都不超出本文中“服务”一词的含义。
以下结合图1从各模块的角度描述本发明实施例的系统100的组成与结构。
图1是本发明的一种面向服务提供者的服务派发系统100的示意图，该系统包括：信息接收模块110、存储模块120、处理模块130以及输出模块140。系统内各模块之间的连接形式可以是有线的，也可以是无线的。任何一个模块都可能是系统本地的，也可以位于远程，通过网络与其他模块连接。模块与模块间的对应关系可以是一对一的，也可以是一对多的集中式，如一个处理模块130同时与多个信息接收模块连接，还可以是多对多的，例如，多个信息接收模块与多个存储模块相互交换信息。
存储模块120和输出模块140不是必须全部存在的，而是可以随着应用场景的变化而有取舍的。例如，对于某些仅依赖实时信息的服务派发系统，可以舍弃存储模块120，而不影响整个系统的运转。又例如，对于某些存在于后台，仅实现信息计算和处理过程的服务派发系统，可以舍弃输出模块140。以上举例仅仅是为了说明存储模块120和输出模块140不是系统所必需的模块，相关领域的技术人员可以根据需要，对上述两个模块的配置进行改进与变化，这些改进与变化都 不脱离本发明的精神与范围。
系统100接收来自一个或多个服务请求者(201-20m)的信息，同时也能接收来自一个或多个服务提供者(301-30n)的信息。特别的，来自服务请求者的信息包括但不限于与其相关的订单信息(211-21m)。此外，系统100还能接收来自信息源400的信息。
服务请求者(201-20m)泛指任何在一定空间范围、时间段内对特定服务发出请求的服务请求方；服务提供者(301-30n)泛指任何在一定空间范围、时间段内提供特定服务的服务提供方。例如，服务请求者(201-20m)可以是路上的需要车辆服务的个人、存在水果需求的个人、需要用餐的个人、需要运送物品的个人或实体等；服务提供者(301-30n)可以是出租车司机、专车司机、水果零售商、餐饮服务提供者、快递员或信使等。
信息源400泛指能向系统100提供信息的来源。信息源400用于存储与服务相关的信息，例如，地理信息、天气情况、交通信息、法律法规信息、新闻事件、生活资讯、生活指南信息、对服务提供者的评价、服务提供者的背景资料、服务请求者的背景资料等。信息源400可以是以一个单独的中央服务器的形式存在，也可以是以多个通过网络连接的服务器形式存在，还可以是以大量的个人设备形式存在。当信息源以大量个人设备形式存在时，这些设备可以通过一种用户生成内容(user-generated content)的方式，例如向云端服务器上传文字、声音、图像、视频等，从而使云端服务器连同与其连接的众多个人设备一起组成信息源。
具体的，如果是交通服务，信息源400可以是包含有地图信息与城市服务信息的市政服务系统、交通实时播报系统、天气播报系统、新闻网络等，或者是某些包含有大量历史订单信息的数据库等。信息源400可以是实物信息源，如常见的测速设备、传感设备，例如司机的车载测速仪、道路上的雷达测速仪、温湿度传感器。信息源400也可以是获取新闻、资讯、道路实时信息等的源，例如一个网络信息源，包括但不限于基于Usenet的互联网新闻组、Internet上的服务器、天 气信息服务器、道路状况信息服务器等。具体的，对于送餐服务，信息源400可以是存储有某一地域众多餐饮服务商的系统、包含有地图信息与城市服务信息的市政服务系统、交通路况系统、天气播报系统、新闻网络等，或者是某些包含有大量餐饮历史订单信息的服务器等。上述举例并非用于局限此处的信息源的范围，也并非局限于所举实例这几类服务范围，本发明可以适用于各种服务任何能够提供与相应服务有关的信息的设备、网络，都可以被归为信息源。
信息接收模块110，接收来自服务请求者(201-20m)、服务提供者(301-30m)、信息源400的信息，将上述信息输入至处理模块。信息收集的方式，包括但不限于，有线或无线方式，向服务提供者、服务请求者获取或询问信息和从信息源400以接收订阅、推送的方式获得信息。
向服务提供者与服务请求者所获取或询问的信息，包括但不限于，涉及服务提供者的信息与涉及服务请求者的信息。以及涉及服务的信息，包括但不限于，服务的具体内容，如出租车服务、个人汽车的搭载、共享用车、快递、送餐、送货、停车、广告宣传、洗衣、维修、娱乐、表演等；服务的提供方式，如上门服务、递送、提供物品的使用权(包括租赁设备或器材)等；服务的支付方式，如现金交易、在线支付、转账汇款等；服务的小费或奖励；服务的时限，如实时服务、预约服务等。
从信息源400以接收订阅、推送的方式所获得的信息，包括但不限于：服务提供者和/或服务请求者的基本资料，如年龄、性别、国籍、住址、族群、宗教信仰、受教育程度、工作经历、婚姻状况、情感状况、语言能力、专业技能、政治倾向、兴趣爱好、喜爱的音乐/电视节目/电影/书籍等；服务提供者和/或服务请求者的生理相关信息，如身高、体重、腰围、胸围、臀围、BMI指数、肺活量、视力、是否色弱/色盲、过往病史、家族遗传病史、病历等；服务提供者和/或服务请求者的历史资料相关信息如驾照信息、交通违规记录、酒驾记录、犯罪记录、信用记录等；服务提供者历史上的服务提供信息，如历史服 务的时间与地点、服务频次、服务内容、接收到的服务评价、服务结果；服务请求者历史上的服务订单信息，如历史服务订单的时间与地点、接收服务的频次、接收服务的内容、接收服务的结果、服务建议或评价；其他信息，如环境参数(如温度、湿度、气压、海拔高度、紫外线强度、风速风向等)、路况(如线路的类型、宽窄程度、吞吐量、车流量、线路拥堵情况、线路上是否发生安全事故或维修施工路面湿滑或结冰等)、天气情况(如降水概率、日照指数、紫外线强度、空气质量指数、细颗粒物/PM2.5、可吸入颗粒物/PM10、森林防火等级、能见度、雨、雪、雾、霾、冰雹等)、自然灾害(如地质灾害指数、台风、龙卷风、泥石流、山体滑坡、地震、火山爆发、海啸等)、地理或地质信息(如时区、气候带、季风带、地质带、地震带、火山群、土壤类型)、关于食物或饮品的信息(如营养事实、热量、来源或产地、加工方法)、文化或族群信息(如某地点所属的文化圈、风俗习惯、民族聚居信息)、信仰信息(如主流宗教信仰、主要宗教场所)、社会活动(如罢工、体育赛事、节日游行、集会、聚众游行、骚乱等)以及其他突发事件等；还可以是飞机航班、火车班次、公共交通班次、船舶班次以及其他交通班次等信息等。
存储模块120，用于存储供处理模块130对信息接收模块100所接收的信息进行运算、判断的判据，并将判据输出到处理模块130。存储模块120，可以包含一个或多个能够存储处理模块130所能访问的信息的任何类型的存储器，包括但不限于常见的各类存储设备如固态存储设备(固态硬盘、固态混合硬盘等)、机械硬盘、USB闪存、记忆棒、存储卡(如CF、SD等)、其他驱动(如CD、DVD、HD DVD、Blu-ray等)、随机存储器(RAM)和只读存储器(ROM)。其中RAM有但不限于：十进计数管、选数管、延迟线存储器、威廉姆斯管、动态随机存储器(DRAM)、静态随机存储器(SRAM)、晶闸管随机存储器(T-RAM)、和零电容随机存储器(Z-RAM)等；ROM又有但不限于：磁泡存储器、磁钮线存储器、薄膜存储器、磁镀线存储器、磁芯内存、磁鼓存储器、光盘驱动器、硬盘、磁带、早期NVRAM(非 易失存储器)、相变化内存、磁阻式随机存储式内存、铁电随机存储内存、非易失SRAM、闪存、电子抹除式可复写只读存储器、可擦除可编程只读存储器、可编程只读存储器、屏蔽式堆读内存、浮动连接门随机存取存储器、纳米随机存储器、赛道内存、可变电阻式内存、和可编程金属化单元等。以上提及的存储设备是列举了一些例子，该网络存储设备可以使用的存储设备并不局限于此。
在某些实施例中，存储模块120，还能用于缓存信息接收模块110与处理模块130双向交互过程中的信息，储存与订单相关的历史信息，供处理模块130使用的其他信息，如历史路况信息、历史天气数据等。
处理模块130，接收来自信息接收模块110的信息，并予以运算处理，并得到一个或多个判断结果。处理模块所执行的运算，可以是基于逻辑的运算，如与或非运算；也可以是基于数值的运算。处理模块可以包含一个或多个处理器，处理器可以是任何通用处理器，例如，一个经过编程的可编程逻辑器件(PLD)，或者一个专用集成电路(ASIC)，或者一个微处理器，也可以是一个系统芯片(SoC)等。
输出模块140，用于接收来自处理模块130运算判断的结果，并将这一结果向服务提供者推送，或者将结果输出至第三方，例如一个信息服务提供商的服务器，由该服务器进行存储，并定期或不定期公开该判断结果，或将该判断结果供服务提供者获取或者查询。推送结果或输出结果的方式，包括但不限于，各种有线或无线通信。
图2显示了根据本发明一个具体实施方式的信息接收模块110的结构图。在此实施方式中，信息接收模块110包括服务请求者信息接收单元111，服务提供者信息接收单元112，外界信息接收单元113。所述三个单元模块，信息流彼此双向联通。其中，服务请求者信息接收单元111接收来自服务请求者20x的信息，服务提供者信息接收单元112接收来自服务提供者30y的信息，外界信息接收单元113接收来自信息源400的信息。
服务请求者信息接收单元111与服务请求者20x之间，服务提供者信息接收单元112与服务提供者30y之间，以及外界信息接收单元 113与信息源400之间的信息获取可以是基于同一种通信方式，也可以是基于不同种类的通信方式。这些通信方式可以是无线的，也可以是有线的。其中无线通信方式包括但不限于IEEE 802.11系列标准、IEEE 802.15系列标准(例如蓝牙技术和紫蜂技术等)、第一代移动通信技术、第二代移动通信技术(例如FDMA、TDMA、SDMA、CDMA、和SSMA等)、通用分组无线服务技术、第三代移动通信技术(例如CDMA2000、WCDMA、TD-SCDMA、和WiMAX等)、第四代移动通信技术(例如TD-LTE和FDD-LTE等)、卫星通信(例如GPS技术等)、和其它运行在ISM频段(例如2.4GHz等)的技术；自由空间光通信包括但不限于可见光、红外线讯号等；声通讯包括但不限于声波、超声波讯号等；电磁感应包括但不限于近场通讯技术等。以上描述的例子仅作为方便说明之用，无线连接的媒介还可以是其它类型，例如，Z-wave技术、其它收费的民用无线电频段和军用无线电频段等。其中有线通信的方式包括但不限于使用金属电缆、光学电缆或者金属和光学的混合电缆，例如：同轴电缆、通信电缆、软性电缆、螺旋电缆、非金属护皮电缆、金属护皮电缆、多芯电缆、双绞线电缆、带状电缆、屏蔽电缆、电信电缆、双股电缆、平行双芯导线、和双绞线。
服务请求者信息接收单元111获取的来自服务请求者20x的信息内容，服务提供者信息接收单元112获取的来自服务提供者30y的信息内容以及外界信息接收单元113获取的来自信息源400的信息内容，已在上文有关描述中涉及，在此不再赘述。
此处，将信息接收模块110划分为三个子模块只是为了说明的方便，显然，对本领域技术人员而言，可以将上述服务请求者信息接收单元111、服务提供者信息接收单元112、外界信息接收单元113中的两者或者全部都集成于一个单独的元件上。例如，将服务请求者信息接收单元111与服务提供者信息接收单元112集成在一个电子元件上，用来收集服务请求者与服务提供者所实时产生的信息。而另外安置一个外界信息接收单元113，用来接收外界信息与历史服务信息。 凡此种种变化与改进，仍应视为本申请所寻求保护的范围。
图3显示了根据本发明一个具体实施方式的处理模块130的结构图。在此实施方式中，处理模块130包括处理器131，控制器132，判据存储单元133。其中，处理器131与控制器132双向交互，判据存储单元133与控制器132双向交互。处理模块130与信息接收模块110和存储模块120双向交互。处理器131可用于基于判据存储单元133存储的判据信息以及来自信息接收模块110或者存储模块120的信息，对特定订单进行判断，并输出至少一个判断结果。控制器132可用于实现对处理器131以及判据存储单元133的控制，以保证处理器正确执行判断功能。判据存储单元133可用于存储于某些判断条件相关的判据信息，并将判据信息用于处理器131的判断过程。所述判据信息可来自于本地，信息接收模块110，也可来自于信息源400。当来自某一服务请求者的服务订单与一个特定的服务提供者的信息被判断满足某一判据时，系统将决定该服务订单可以被派发至该服务提供者；如果该信息被判断不满足某一判据时，系统将决定该服务订单不被派发至该服务提供者。以下将对判据存储单元133进行进一步介绍。
以上所说的处理器、控制器与判据存储单元仅是就它们所各自执行的功能而说的，在实际应用中，可以将同一个硬件设备，如一个专用集成电路，或一个微处理器，或者其他任何可以执行计算处理信息并进行逻辑判断的器件或设备，配置为能够执行处理器、控制器与判据存储单元的全部功能，或配置为能够执行处理器、控制器与判据存储单元中任意两者的功能，而不脱离本发明所寻求保护的范围。
图4显示了根据本发明一个具体实施方式的判据存储单元133的结构图。在本实施方式中，判据存储单元133包含服务提供者服务范围判据1331、服务提供者活跃程度判据1332、抢单概率判据1333、路线障碍物判据1334、订单顺路判据1335以及其他判据1336。所述判据对应不同判断模式，分别对应服务范围判断、服务提供者活跃程度判断、抢单概率判断、路线障碍物判断、顺路订单判断以及其他判 断模式。该判断过程在控制器132中完成。
图5a是服务派发系统流程图。本实施方式中，在步骤S100，由信息接收模块110接收来自服务请求者20x的订单信息；在步骤S110，由信息接收模块110接收来自服务提供者30y的司机当前信息；在步骤S120，由处理模块130，将当前订单信息与服务提供者当前信息进行计算分析，得到一个特征结果；在步骤S130，由处理模块130，基于判据存储单元133中至少一个判据，进行判断：该特征结果是否满足预设定的至少一个判据，若结果为是，则进入步骤S140，向该服务提供者派发该订单；若结果为否，则进入步骤S150，不向该服务提供者派发该订单。
判断流程也可以根据多个判据，不同判据之间可以采取并行或者串行的结构关系。在串行的结构关系中，即使只有一个判据的判断结果为否，则流程结束，不向该服务提供者派发订单。在并行的结构关系中，只要有一个判据的判断结果为是，则流程结束，向该服务提供者派发订单。
另外，判断流程也可以是一个基于多个判据的综合判断打分流程，通过给不同判据分配各自的权重，得到一个综合分数，以这一分数与一个预设的阈值进行比较，以决定判断结果。
判断流程还可以是一个根据整个系统结果进行的最优化选择过程，在确定了多个服务提供者作为候选服务提供者后，进一步选择判据，继续判断/打分过程，进一步筛选后续服务提供者，直到最终挑选出最优化的一个或多个服务提供者。
具体地，该特征结果是根据服务请求信息、来自服务请求者与服务提供者的信息产生的。产生后，该特征结果被处理模块用来与判据存储单元所存储的判据进行比较判断。判断的方式包括，但不限于：将特征结果中的一个特征值与判据中的一个阈值进行对比；判断一个特征值是否落入判据中的一个数值范围；将一个特征值可以按照判据中的排名/给分规则(ranking rule/rating rule)进行排序或分配分数，并与一个次序或分数的阈值进行比较；判断特征结果是否触发判据中 的某一事件；判断特征结果是否符合判据中的某一模式等。
图5b是结合了外界信息的服务派发系统流程图。本实施方式中，在步骤S105，由信息接收模块110接收来自服务请求者20x的订单信息；在步骤S115，由信息接收模块110接收来自服务提供者30y的服务提供者当前信息；在步骤S125，信息接收模块还接收来自外界信息源400所提供的信息，在步骤S135，由处理模块130，将当前订单信息、服务提供者当前信息以及外界信息进行计算分析，得到一个特征结果；在步骤S145，由处理模块130，基于判据存储单元133中至少一个判据，进行判断：该特征结果是否满足预设定的至少一个判据，若结果为是，则进入步骤S155，向该服务提供者派发该订单；若结果为否，则进入步骤S165，不向该服务提供者派发该订单。
服务派发系统可以充分获取服务请求者的信息，并结合服务提供者的信息进行匹配。
这里，服务请求者的信息，可以包括服务请求者的定位与运动信息、当前生理/健康状态、当前心理状态、对服务提供者的偏好、对服务形式或内容的偏好等信息。
其中，定位与运动信息，可以包括但不限于，当前位置、当前运动状态、当前运动方向、当前运动速度、当前活动状态等；当前生理/健康状态，可以包括但不限于，饥饿、饱腹、疾病、血压、脉搏、心律、体温、心电图、脑电波、呼吸频率、血糖含量、血氧含量等；对服务提供者的偏好，可以包括但不限于，在个人信息、从业经验、其他能力或特长、信仰或政治倾向上的预期或偏好；对服务形式或内容的偏好，可以包括但不限于，服务支付方式、送货上门、送餐上门、运送速度、接驾速度等信息。
这里，服务提供者的信息，可以包括服务提供者的定位与运动信息、当前生理/健康状态、当前心理状态、运营信息、对服务请求者的偏好等信息。
其中，定位与运动信息，可以包括但不限于，当前位置、当前运动状态、当前运动方向、当前运动速度、当前活动状态等；当前生理 /健康状态，可以包括但不限于，饥饿、饱腹、疾病、血压、脉搏、心律、体温、心电图、脑电波、呼吸频率、血糖含量、血氧含量等；运营信息，可以包括但不限于，当前运营状态、当前所提供的服务内容、服务时间、服务地域范围、服务常驻点、服务对象、服务方式；对服务请求者的偏好，可以包括但不限于，对服务请求者在个人信息、文化/教育/职业信息、技能特长等方面的偏好等。
在匹配的过程中，所利用的服务请求者或服务提供者的信息，可以是一项，也可以是多项。
服务请求者信息的挖掘
具体地，服务请求者的信息，可以包括服务请求者的定位与运动信息、当前生理/健康状态、当前心理状态、对服务提供者的偏好、对服务形式或内容的偏好等信息。
其中，定位与运动信息，可以包括但不限于，当前位置、当前运动状态、当前运动方向、当前运动速度、当前活动状态等；当前生理/健康状态，可以包括但不限于，饥饿、饱腹、疾病、血压、脉搏、心律、体温、心电图、脑电波、呼吸频率、血糖含量、血氧含量等；对服务提供者的偏好，可以包括但不限于，在个人信息、从业经验、其他能力或特长、信仰或政治倾向上的预期或偏好；对服务形式或内容的偏好，可以包括但不限于，服务支付方式、送货上门、送餐上门、运送速度、接驾速度等信息。
上述信息的采集，可以通过多种智能设备或者专用的测量设备来完成，例如，与位置或运动有关的信息，可以通过带有定位功能的设备来采集；与健康、生理有关的信息，可以通过带有多种传感器的智能穿戴设备或医用设备来采集；对于心理状态、对服务提供者的预期以及对服务形式或内容的预期，可以通过服务请求者在智能设备上以文字、语音等形式输入而采集。
对服务请求者服务偏好的挖掘
通常，服务请求者在寻求服务时，会存在对服务提供者、服务内容或服务形式的偏好。在某些场合，服务内容本身不是唯一的，而具 有某种可替换性，但服务请求者对于一类仅仅在形式上存在差别的同质化服务，仍然存在某些个人倾向。挖掘这些倾向，有助于实现效率更高、更有针对性、更具有个性化的服务派发。
例如，对于寻求交通服务的服务请求者，其偏好可能包括但不限于，对于服务提供者(司机)在个人信息、从业经验、能力或特长、信仰或政治倾向上的要求或偏好。其中，个人信息，可以包括但不限于，性别、年龄、婚姻状况、情感状况、受教育程度等。从业经验，可以包括但不限于司机的驾龄、驾照类型、通过驾照考试时的成绩排位、驾驶过的车型、交通事故记录、交通违规记录、乘客评价等。能力或特长，可以包括但不限于，兴趣爱好、语言能力、运动特长等。信仰与政治倾向，可以包括但不限于，司机的宗教信仰、政治倾向、党派信息、所参加的社会团体等信息。除此之外，对于寻求交通服务的服务请求者，还可能存在对于服务内容或形式上的偏好或倾向，这些包括但不限于，对于搭载的车辆性能上的要求或偏好、对于交通服务内容的要求或偏好。对搭载的车辆的性能上的要求或偏好，可以进一步包括但不限于，对车辆的品牌、型号、最高车速、加速时间、耗油量、马力、最大加速度、排量、排放标准等方面的要求或偏好。对交通服务内容的要求或偏好，可能包括但不限于，对于司机接驾的要求、对交通服务体验的要求等。优选地，乘客对于服务提供者的偏好可以包含有服务提供者的从业经验与能力或特长信息，进一步优选地，从业经验与能力或特长，可以包括服务提供者的驾龄和语言能力，例如，某位来到柏林的巴西游客可能希望司机具备葡萄牙语的基本听说能力。
又例如，对于寻求送餐服务的服务请求者，其偏好可能包括但不限于，对于服务提供者(餐馆和/或送餐人员)的偏好以及对于餐饮本身的偏好。对于服务提供者(餐馆和/或送餐人员)的偏好，可以包括但不限于，对餐馆的营业经历(包括但不限于，开业时间、营业范围、餐饮卫生状况、餐饮卫生事故、顾客评价)的偏好、对送餐人员送餐速度的偏好、对送餐人员的从业经验或业内评价的偏好。对于餐饮本 身的偏好，可以包括但不限于，对餐饮品种的要求、对饮食口味(如酸、甜、苦、咸、鲜等味觉)的偏好、对于餐食营养价值的偏好(如热量值、包括碳水化合物、脂肪、蛋白质、矿物元素、维生素在内的主要营养物质的指标)、食物安全等级(如转基因食品、有机食品、绿色无公害食品等)上的要求或偏好等。
又例如，对于寻求洗衣服务的服务请求者，其偏好包括对服务提供者的偏好与对服务内容的偏好。其中，服务请求者对服务提供者的偏好可能包括但不限于，服务提供者的从业情况(如从业时间、从业经验、从业资质)、服务提供者的顾客评价等。对服务内容的偏好可能包括但不限于，洗衣的方法(如手洗、干洗、机洗等)、脱水方法(如离心脱水、机器烘干、阳光晾晒等)等的偏好。
以上举例仅仅是为了说明目的，并非用于限制本文中偏好的含义。事实上，任何与服务请求者对服务的倾向有关的信息，都属于本文中偏好的范围。
挖掘服务请求者的偏好，可以利用该服务请求者的历史服务信息来实现。历史服务信息中包含有当时服务提供者的个人信息、从业经验、能力或特长、信仰或政治倾向等信息，还可能包含服务形式或内容的具体信息。这些信息，都可以被处理模块获取并分析。
例如，对于历史交通服务，这些信息可能包括但不限于，服务提供者(司机)的个人信息、从业经验、能力或特长、信仰或政治倾向。其中，个人信息，可以包括但不限于，性别、年龄、婚姻状况、情感状况、受教育程度等。从业经验，可以包括但不限于司机的驾龄、驾照类型、通过驾照考试时的成绩排位、驾驶过的车型、交通事故记录、交通违规记录、乘客评价等。能力或特长，可以包括但不限于，兴趣爱好、语言能力、运动特长等。信仰与政治倾向，可以包括但不限于，司机的宗教信仰、政治倾向、党派信息、所参加的社会团体等信息。除此之外，历史交通服务还可能存在服务内容或形式上的信息，这些包括但不限于，搭载的车辆性能信息、交通服务内容的信息。搭载的车辆的性能信息，可以进一步包括但不限于，车辆的品牌、型号、最 高车速、加速时间、耗油量、马力、最大加速度、排量、排放标准等方面信息。交通服务内容的信息，可能包括但不限于，司机接驾信息、交通服务体验等。
又例如，历史送餐服务中的信息，可能包括但不限于，服务提供者(餐馆和/或送餐人员)的信息以及餐饮本身的信息。服务提供者(餐馆和/或送餐人员)的信息，可以包括但不限于，餐馆的营业经历(包括但不限于，开业时间、营业范围、餐饮卫生状况、餐饮卫生事故、顾客评价)、送餐人员送餐速度、送餐人员的从业经验或业内评价。餐饮本身信息，可以包括但不限于，餐饮品种、饮食口味(如酸、甜、苦、咸、鲜等味觉)、餐食营养价值(如热量值、包括碳水化合物、脂肪、蛋白质、矿物元素、维生素在内的主要营养物质的指标)、食物安全等级，例如转基因食品、有机食品、绿色无公害食品等。
又例如，历史洗衣服务中的信息，可以包括但不限于，服务提供者的信息与服务内容的信息。其中，服务提供者的信息可能包括但不限于，服务提供者的从业情况(从业时间、从业经验、从业资质)、服务提供者的顾客评价等。服务内容信息可能包括但不限于，洗衣的方法(手洗、干洗、机洗等)、脱水方法(离心脱水、机器烘干、阳光晾晒等)。
例如，根据某一个人此前的历史订餐信息，可以挖掘该个人可能倾向的餐饮品种、饮食口味、餐饮价位、送餐速度等。具体地，可以抓取该个人在一段时间(例如，一年、半年、一月)的订餐记录，计算历史订餐总次数，并对每一次订餐记录信息中所包含的餐饮品种、饮食口味、餐饮价位、送餐速度进行统计，据此可以得出该个人的最可能倾向的餐饮品种、饮食口味、餐饮价位、送餐速度。
在某些情形下，尽管偏好有明显的个体差异，但也呈现出群体特征。可以通过对服务请求者的人口统计学信息、背景资料等信息进行归类、分析，从而预测某一服务请求者的可能偏好。
例如，可以根据历史服务订单数据中同一类人群的主流或共同偏好，预测当前服务请求者的可能偏好。该类人的划分，可以是基于多 种不同的参数或属性，包括但不限于，个人信息(年龄、性别、国籍、家乡、现住址)、文化/教育/职业信息(受教育程度、就读高校、毕业高校、文化圈、职业、就职单位)、生理信息(健康情况、体形、血型、身高、体重)、信仰与政治倾向(宗教信仰、政治倾向)等。
优选地，可以根据个人信息来预测当前服务请求者的可能偏好。进一步优选地，该个人信息，可以包括，性别、年龄、国籍、家乡。
另一个优选的情况，可以根据教育或职业信息，来预测当前服务请求者的可能偏好。进一步优选地，该文化/教育/职业信息，可以包括，受教育程度、学历、学位、就读/毕业的高中、就读/毕业的高校、所学专业、所属文化圈、职业、所属行业、职业水平、就业单位、从业时间等。
另一个优选的情况，可以根据生理信息，来预测当前服务请求者的可能偏好。进一步优选地，该生理信息，可以包括，健康状况、身高、体重、体型、视力、血型等。
再一个优选的情况，可以根据信仰与政治倾向，来预测当前服务请求者的可能偏好。进一步优选地，该信仰或政治倾向，可以包括，信奉宗教、宗教派别、信教时间、教内职务、政治党派、支持的政治阵营。
以上信息可以结合，共同判断或预测某个体的可能偏好。例如，可以将个人信息与文化/教育/职业信息予以综合考虑，以准确地预测偏好。
例如，若通过挖掘历史服务订单信息，发现在纽约市的中国留学生发起的餐饮订单中的餐食大多是川菜与粤菜，那么，一旦发现送餐服务请求者来自重庆，现在纽约城市大学攻读博士学位，那么，服务派发系统在分析来自信息模块的关于该请求者的个人信息后，就会优先将他/她的送餐服务请求推送到纽约市内的川菜馆或粤菜馆。
服务提供者信息的挖掘
具体地，服务提供者的信息，可以包括服务提供者的定位与运动信息、当前生理/健康状态、当前心理状态、运营信息、对服务请求者 的偏好等信息。
其中，定位与运动信息，可以包括但不限于，当前位置、当前运动状态、当前运动方向、当前运动速度、当前活动状态等；当前生理/健康状态，可以包括但不限于，饥饿、饱腹、疾病、血压、脉搏、心律、体温、心电图、脑电波、呼吸频率、血糖含量、血氧含量等；运营信息，可以包括但不限于，当前运营状态、当前所提供的服务内容、服务时间、服务地域范围、服务常驻点、服务对象、服务方式；对服务请求者的偏好，可以包括但不限于，对服务请求者在个人信息、文化/教育/职业信息、技能特长等方面的偏好等。
优选地，来自服务提供者的信息，包括服务提供者的实时信息，更进一步优选地，该服务提供者的实时信息，包括当前位置、当前运动状态、当前运动速度与当前运营状态。
优选地，来自服务提供者的信息，包括服务提供者的当前生理状态，更进一步优选地，该服务提供者的当前生理状态，包括当前心律、血压、体温、心电图、脑电波、血糖含量、血氧含量等。
上述信息的采集，可以通过多种智能设备或者专用的测量设备来完成，例如，与位置或运动有关的信息，可以通过带有定位功能的设备来采集；与健康、生理有关的信息，可以通过带有多种传感器的智能穿戴设备或医用设备来采集；对于心理状态、当前运营状态、以及对服务请求者的偏好，可以通过服务提供者在智能设备上以文字、语音、图形等形式输入而采集，也可以通过对历史数据的分析而获取。
优选地，可以通过带有定位功能的设备，采集服务提供者的当前位置、运动状态与运动速度信息。更进一步优选地，该设备，包括但不限于，智能手机、平板电脑、车载定位设备、导航仪等。再进一步优选地，该设备采用一种定位技术获取当前位置、运动状态与运动速度信息，该定位技术，选自全球定位系统(GPS)技术、全球导航卫星系统(GLONASS)技术、北斗导航系统技术、伽利略定位系统(Galileo)技术、准天顶卫星系统(QAZZ)技术、基站定位技术、Wi-Fi定位技术、交通工具自带的各种定位测速系统等。
优选地，可以通过内置有传感器的设备，采集服务提供者的生理状态。进一步优选地，该设备上内置的传感器包括但不限于光电传感器、血压传感器、心电信号传感器、体温传感器、热通量传感器、脉搏波传感器、生物电传感器、三维运动传感器等。更进一步优选地，该设备，包括但不限于，智能可穿戴设备、心电仪、脑电波测量设备。再进一步优选地，智能可穿戴设备包括但不限于，智能手表、智能手环、智能头盔、智能衣物、智能眼镜、健康贴片。
服务范围、常驻点(似然地理位置)的确定
对于服务提供者，服务范围是一个重要概念。不失一般性，服务范围，通常是指服务提供者愿意执行服务内容的地理区域。不同行业的服务提供者，可能会有不同的服务范围，例如，出租车司机通常会将某个城市的主要城区作为服务范围；提供外卖服务的餐饮服务商，可能仅在某几个街区提供送餐服务，相应的，所述的几个街区就构成了该餐饮服务商的服务范围。
对于与位置相关的服务，常驻点，或似然地理位置，是另外一个重要概念。常驻点，是指某一个人或实体所频繁出现的地点，或者该个人或实体长期所处的位置。例如，对于车辆司机，常驻点可能是该司机的家庭住址、中午就餐地点；对于餐饮提供商或家政服务提供者，常驻点可能是指该商户或该个人的营业场所。
服务提供者的服务范围与常驻点，可以通过对历史服务信息的挖掘而获得。如果历史服务信息中包含有服务的地点信息，例如，交通服务的起点/终点、送餐服务的起点/目的地、家政服务的服务人员的住址/营业场所/服务地点等。通过对大量历史服务信息中的地点信息的收集，服务派发系统可以获得服务提供者的服务范围/常驻点信息。
以下通过实例，描述挖掘常驻点信息的方法与过程。
图6是获得服务提供者常驻点的流程图。
在本实施例中，服务提供者常驻点信息，既可以是来自服务提供者的输入，如服务提供者在系统中所关联的账户中，记录自己的家庭住址、工作地址、经常停留的地点；也可以是由一个计算设备对服务 提供者历史位置的分析、挖掘所得到的常驻点信息。
本实施例中，对常驻点的挖掘可以基于来自服务提供者历史位置与历史轨迹。具体地，对常驻点的挖掘可以包含以下步骤：
步骤S200，在预设的时间周期内获取服务提供者轨迹信息，形成一服务提供者信息集合；
时间周期可以为一周、一个月、一个季度、半年、一年，但也可以更长或者更短，具体的时间周期在此不做限定，一般时间越长，计算越准确，但计算量与所需要的存储空间也越大；
步骤S210，根据服务提供者信息集合，计算出该服务提供者在上述时间周期内的常驻点。
服务提供者信息集合，是关于服务提供者的多项参数的信息集合，其中的多项参数包括但不限于：服务提供者身份(ID)、报告信息的时间(时间戳)、服务提供者所处位置(包含有经度、纬度)，以及，可选地，在所处位置的停留时间。信息集合还可以包含其他的参数，比如，服务提供者在所处位置的活动内容等。
在某些情况下，服务提供者在某一位置所停留时间并不需要显式地提供给系统，系统可以根据多条司机信息记录，计算出在该位置的停留时间。
根据服务提供者信息集合，计算服务提供者的常驻点，可以首先得到多个候选常驻点，然后对这些候选常驻点进行分析，获得最终的候选点信息。
在得到候选常驻点与最终的常驻点时，可以采用聚类算法。所述聚类算法自定义一段距离和一段时间，根据服务提供者信息集合中的经纬度和停留时间自动将该距离和时间内的经纬度归为某一服务提供者的常驻密度区域。
在一个例子中，可以采用Dbscan算法计算服务提供者常驻点，然而，可以理解，用于计算服务提供者常驻点的算法，不限于Dbscan算法，还可以有其他方法，如划分方法(Partitioning method)，如K-means、K-medoids、CLARA(Clustering LARge Application)、 CLARANS(Clustering Large Application based upon RANdomized Search)、FCM等；层次法(hierarchical method)，如BIRCH(Balanced Iterative Reducing And Clustering using Hierarchies)、正二进制(binary-positive)方法、连续数据的粗聚类算法(rough clustering of Sequential data，简称RCOSD)；基于密度的方法，如OPTICS(Ordering Points To Identify The Clustering Structure)；基于网络的方法，如STING(STatistical INformation Grid)、CLIQUE(Clustering In QUEst)、Wave-Cluster；基于模型的方法，如Cobweb、CLASSIT等。
以交通服务为例，图7是利用处理模块判断司机常驻点的流程图。在步骤S300，预设一时间周期T，设在所述时间周期T内获取某一司机的轨迹信息，并形成一司机信息集合，在所述司机信息集合中包含有n个经纬度坐标，将每个经纬度坐标作为司机的候选常驻点A1、A2、…、An；
在某些情况下，智能设备可能不会随时获知司机的位置并上传位置数据，而是只在司机开启智能设备的定位功能或关闭定位功能时执行上述操作。在这些情况下，候选常驻点可能是每天司机的智能设备所上传的第一个地点和/或最后一个地点。例如，如果司机每天离开家时，打开智能设备的定位功能，回到家时，关闭该功能，则第一个地点与最后一个地点重合，即为家庭住址。不失一般性，在上述情况下，每天可能会有两个候选常驻点被记录。有的司机一天可能会在一段时间内关闭定位功能，然后再开启定位功能，这样一天可能就有四个候选常驻点被记录，每天候选常驻点的记录数量以司机开启和关闭定位功能时上传的经纬度数量为准，按照此方法算出来的常驻点，一般是司机的家庭住址，因为司机一般在离开家或者到家的时候才会开启或者关闭定位功能；
在另一些情况下，智能设备可能每隔一段时间获知司机的位置并上传位置数据，这段时间可能是一个预先设定的时间间隔，也可能是一种随机的时间间隔。候选常驻点的选择并不一定是司机开启或者关闭定位功能时上传的经纬度，而是将定位功能中每隔一段时间所自动 上传的所有地点作为候选常驻点，然后将停留在候选常驻点的时间大于一阈值的点留下，因为这样算出的常驻点有可能是司机午饭停留的地点或者其他地点，候选常驻点的设定方式在此不做具体限定。
在步骤S310，计算每个候选常驻点到其他候选点的总距离，删除总距离最长的候选常驻点，当剩余的候选常驻点数量小于0.5n时，执行步骤S330，否则重复步骤S320；
在步骤S330，如果任意两个候选常驻点之间的距离大于一距离阈值，则进入步骤S350，认为司机无常驻点，否则执行步骤S340；
步骤S300到步骤S330为常驻密度区域的计算步骤。步骤S320到步骤S330可以理解为一种去噪的方法，对于所有的常驻点候选点，计算其到其他点的总距离。总距离最大的点，可以作为一个噪声点去除，因为这个点距离可能的中心点最远，循环的按照这种方法去噪，保留的点更趋近于司机的常驻点，该循环可以通过如下方法判断是否结束：即剩余的所有候选点，两两之间的最大距离不超过一个定义的阈值，一般这个阈值为1公里到5公里。于是，假如有N个常驻点候选点，通过这种循环的方法，可以过滤掉1/4到3/4的噪声点。
图8是处理模块判断常驻点的去噪示意图。图上方的越靠上的点越早被过滤，501、502、503、504、505与506这6个点会先后被作为噪声点去除。
回到图7，在步骤S340，利用剩余的所述候选常驻点，计算出常驻点。
这里计算常驻点的方法，可以是将所述候选常驻点的经纬度数值，进行平均值的求解。这里求解平均值的方法，可以是算术平均值，也可以是几何平均值。
在所述S300中，所述候选常驻点的设定可以根据在所述候选常驻点的停留时间；
在所述S340之前，还可以通过设定时间阈值来进一步筛选候选常驻点，在一些情况下，可以设定一个时间阈值，如20分钟，如果司机在某一候选常驻点所停留的时间始终少于20分钟，则删除该候 选常驻点。
具体的，可以先删除在所述候选常驻点的停留时间小于一时间阈值的候选常驻点，将剩余的所述候选常驻点通过求几何平均值的方法计算出常驻点。
用于判定常驻点的时间阈值，对于不同司机、不同车辆而有不同，例如但不限于，15分钟、30分钟或者更长。具体的常驻点判断，需要结合司机个人、车辆情况、时间、地点、路况等诸多因素综合考虑。例如停留时间在5分钟之内的点可能是司机遭遇堵车时所处于的地点，并不能作为计算常驻点的数据。
在获取了常驻点信息之后，可以将这一信息用于服务订单的派发判断中。例如可以将服务提供者的常驻点信息与服务订单的位置信息结合，以供处理模块进行判断。如果某一服务订单中包含的一个地理位置位于某一服务提供者的常驻点附近，该订单可能对于该服务提供者而言，是较为方便执行的订单，因此可以将该订单派发给该服务提供者。更深入的细节，可以参见后文中的关于顺风车判断的描述。
虽然，以上实施例描述了挖掘司机的常驻点信息的方法与流程，但是，需要注意的是，能够被挖掘常驻点信息的服务提供者，并不仅限于司机，任何其他服务提供者，例如，提供餐饮服务的商户、提供送货服务的快递员、送信人、提供洗衣服务的个人或商户、提供家政服务的个人或商户等，都可能存在似然地理位置或常驻点，相应地，他们的常驻点信息都可能通过类似的方法或流程，被服务派发系统所获取。
营业时间(opening time)、忙碌时段(busy period)的确定
从信息源400，系统可以获知历史服务信息，通过对这些服务信息的数据挖掘，从中进一步获得服务提供商的营业时间、忙碌时段信息。
营业时间，又称服务时段，是指服务提供商提供、执行服务的时间。在这一个时间段内，服务提供商可以对服务请求者的服务请求做出应答，并执行服务内容。
忙碌时段，是指在服务提供商的营业时间内，服务提供商的执行服务的频率或负载达到一定阈值，难以应答潜在的服务请求的一段时间。例如，尽管某家餐饮提供商在平日7:00-23:00营业，但只有每天早上7:30-9:00、中午11:00-13:00、傍晚17:00-18:30以及晚上20:00-22:00才处于忙碌状态，相应地，这四段时间构成了该餐饮商户的忙碌时段。
服务提供商，通常装备有专门的服务订单管理的计算机、服务器，用来存储历史服务订单数据。服务派发系统，可以根据从信息源400所获取的历史服务订单数据，利用大数据挖掘、模式分析等手段，获知特定的服务提供商者的营业时间、忙碌时段等信息。
具体地，对于某一服务提供者，可以收集一定时间段的历史服务订单信息，将这一时间段按照一定时间间隔划分成多个子时段，统计每一个子时段内的服务订单数量，比较不同子时段内的服务订单数量，即可以识别出营业时间与忙碌时段。
参考图9，对获得服务提供者的忙碌时段进行说明。
首先，在步骤S400中，服务派发系统通过历史服务信息数据库，调取一个预定时间期间T内，某一服务提供者的所有服务信息，共计N条，以此构成一个服务信息集合；
在步骤S410，将这个期间T划分为m个子时段，记为t1、t2、…、tm；
进入步骤S420，对每一个子时段内的服务信息进行计数计数结果为n1、n2、…、nm；
在步骤S430，通过把不同的ni(1≤i≤m)进行比较，找到ni中的最大值nk，并查找到相应的tk；
最后，在步骤S440，将tk作为服务提供者的一个忙碌时段。
需要注意，预定时间期间T，可以随着服务内容的不同、服务业整体运营状态不同而有所差异，可以是一周、一个月、一个季度甚至一年等。子时段的划分也与此类似，子时段的长度可以为半小时、一小时、两小时、三小时、十二小时、一天、一周等。通常子时段的长 度要短于整个时间段的长度。子时段t1、t2、…、tm的划分，并不必然是均匀的，不同子时段的长度可以相同，也可以不同，如对7:00-9:00时间段，可以划分为7:00-7:30、7:30-7:45、7:45-8:00、8:00-8:15、8:15-8:30、8:30-9:00六个子时段。
以交通服务作为一个例子，优选地，可以设定时间T为一天，每一个子时段可以都设置为半小时，考察在一天之内每半小时时间段内的司机营业情况，计算出最忙碌的时段；另外一种优选的情况，设置一周作为T的长度，每一个子时段可以都设置为一天，考察一周之内每天司机的营业情况，以计算出最忙碌的一天。
以洗衣服务作为另一个例子，优选地，可以设定时间T为一年，每一个子时段可以都设置为一个季度，考察在一年之内每季度内的洗衣服务情况，计算出最忙碌的季节；另外一种优选的情况，设置一年作为T的长度，每一个子时段可以都设置为一个月，考察一年之内每月的营业情况，以计算出最忙碌的一月。
另外，尽管本示例中将服务数量最多的时段作为忙碌时段，但这种标准并非唯一的，可以在子时段中，将订单数量前三名或者前五名的子时段，作为三个或者五个忙碌时段。
对服务提供者的服务内容特点的挖掘
通常，服务提供者所提供的服务在形式、内容方面有一定的特点。例如，对于提供交通服务的服务提供者，其服务特点可能包括但不限于，平均行驶速度快、车内空调温度低、守时等；对于提供送餐服务的服务提供者，其服务特点可能包括但不限于，所提供的餐饮品种、餐饮口味(如酸、甜、苦、咸、鲜等味觉)、食物加工方法、食物新鲜程度、平均送餐时间、送餐方式；对于提供洗衣服务的服务提供者，其偏好可能包括但不限于，洗衣的方法(手洗、干洗、机洗)、脱水方法(离心脱水、机器烘干、阳光晾晒)、平均洗衣时间、递送衣物的速度等特点。
以上举例仅仅是为了说明目的，并非用于限制本文中偏好的含义。事实上，任何与服务提供者提供服务的特点有关的信息，都属于 本文中偏好的范围。
可以通过对历史服务信息的处理，挖掘某一服务提供者提供的服务的特点，从而能够为该服务提供者推荐、派发更有针对性的服务请求信息。
挖掘服务提供者的服务内容的特点，可以利用该服务提供者的历史服务信息来实现。例如，根据某一餐饮商户此前的历史送餐信息，可以挖掘该餐饮商户的主要餐饮品种、饮食口味、餐饮价位、送餐速度等。具体地，可以抓取该餐饮商户在一段时间(例如，一年、半年、一月)的送餐记录，计算历史送餐总次数，并对每一次订餐记录信息中所包含的餐饮品种、饮食口味、餐饮价位、送餐速度、服务评价进行统计，据此可以得出该餐饮商户的最可能倾向的餐饮品种、饮食口味、餐饮价位、送餐速度、服务评价。
服务提供者活跃程度的获知
服务提供者在提供服务时的用户体验与服务请求者接受服务时的用户体验不同。服务提供者的用户体验主要就是服务订单的数量，能否接收到服务订单或者接收到服务订单的数量。用户体验较差的服务提供者，往往难以收到服务订单，因此其收益与服务积极性也较低。因此，判断服务提供者的活跃状况，并对非活跃服务提供者进行针对性策略运营，从而挽留住这些之前用户体验较差的服务提供者，对于提升整个服务派发系统的运作效率是十分必要的。
获取服务提供者的服务活跃程度，可以利用历史服务信息而实现。
服务派发系统，可以接收到来自服务提供者的信息，也可以接收到关于服务提供者的历史信息，统计信息。其中，历史信息与统计信息包括历史在线情况、历史抢单情况、历史接单情况、历史服务信息等。
基于接收到的信息，系统可以根据各服务提供者一段时间内的在线情况和/或抢单情况，发现出某一服务提供者是否为非活跃服务提供者。这种服务提供者可以是餐饮配送服务中的配送人员、交通服务中 的出租车/私家车司机、到家美容服务中的上门服务人员等。
服务提供者的在线情况和/或抢单情况，可以通过被服务提供者安装的服务应用所获得，这些服务应用包括但不限于，安装在个人智能终端上的交通服务app、安装在计算机上的订餐系统、多种服务应用的客户端。
如果通过在线状态，发现某一服务提供者在最近一段时间内，在线时间较短、上线频率较低、未对服务请求信息作出应答、应答时间长、中断服务频率高等情况，服务派发系统可能将该服务提供者列入不活跃服务提供者的名单中。
服务提供者抢单概率的预估
服务提供者的服务记录和抢单日志可以被存储在服务器上，作为服务提供者的历史数据的一部分。运用机器学习和数据挖掘的方法，利用这些海量的历史数据对模型进行训练，可以应用模型从而精准地预估出关于服务提供者选择每次呈现的订单的抢单概率。将预估出来的服务提供者抢单概率用于后期服务派发系统对服务订单的分配。
服务派发系统，可以接收到来自信息接收模块的服务提供者的当前信息和/或历史信息和/或统计信息。其中，历史信息与统计信息包括历史在线情况，历史抢单情况，历史订单相关信息。基于接收到的信息，系统可以根据各服务提供者一段时间内的在线情况和/或抢单情况，判断出关于服务提供者选择每次呈现的订单的抢单概率。这种服务提供者可以是餐饮配送服务中的配送服务人员，同城送货服务中的送货人员，到家美容服务中的上门服务人员等。
这种处理订单的方法，可以由以下步骤实现：获取历史订单的至少一个特征以及与历史订单相关联的服务请求者的响应(是否抢单、是否抢单成功)；根据与历史订单相关联的服务请求者的响应，向至少一个特征分配权重；获取当前订单的特征；以及根据与当前订单的特征相对应的权重，选择当前订单中的将向服务请求者呈现的当前订单。在食物配送服务场景中，这种特征可以包括送餐距离，送餐金额大小，配送费等。
在某些实施方式中，机器学习模型可以是广泛运用于二分类问题的逻辑回归(logistic regression)模型；在某些实施方式中，还可以是支持向量机(support vector machines)模型；在其它实施方式中，还可以根据测试结果来选择使用其它的机器学习模型。
当前订单是指有待向服务请求者呈现或者正在呈现的订单。例如，当前订单可以是尚未向服务请求者呈现的订单，或者是正在向一些服务请求者呈现而尚未向其他服务请求者呈现的订单。可以由在线服务器中获取当前订单。例如在车辆服务场景中，获取订单的方式可以包括直接从发出订单的待乘坐出租车的人员接收订单或者接收由其他中间机构(例如，某个网站服务器等)转发的订单。
在获取当前订单之后，可以在众多服务请求者中选择与当前订单相关联的一个或多个服务请求者作为将向其呈现当前订单的候选服务请求者。作为示例，可以选择处于发送当前订单的位置某范围内的服务请求者作为候选服务请求者。还可以根据利用其它因素、例如服务请求者的行驶方向等来选择候选服务请求者。另外，还可以对所选择的候选服务请求者进行进一步的过滤。
注意到，在获取多个当前订单之后，可以有多个当前订单与单个服务请求者相关联，例如在该服务请求者处于多个当前订单的发送位置的特定范围内的情况下。因而，可以选择多个当前订单中的一个优选订单向该服务请求者呈现。当前订单的特征也可以包括：发送订单的位置与服务请求者的距离或者待乘坐出租车的人员等待出租车时所在的位置与服务请求者的距离、订单中将要前往的目的地，订单的目的地种类(例如，机场、医院或者学校)、订单的目的地周围的路况或者订单的呈现次数。当前订单和服务请求者的特征还可以包括：愿意支付的额外小费、愿意等待的时间、乘坐人数、是否携带大件行李等。此外，如关于历史订单进行描述的，当前订单的特征可以是从当前订单所确定的内容直接确定的，或者可以是利用服务器对所确定的内容进行处理而进一步间接确定的。针对当前订单的特征，可以利用与当前订单的特征相对应的权重，即在训练阶段向与当前订单的特 征相对应的特征分配的权重
对于某一类交通服务，可以指定交通服务的距离参数作为服务订单的特征，分析某一服务提供者对不同距离交通服务订单的响应程度，训练服务派发系统，并对于一个实时服务订单，预估该服务提供者的响应程度。该方法包括：获取当前订单的始发地与目的地之间的距离；获取该服务请求者对于历史订单的抢单概率，其中该历史订单的始发地与目的地之间的距离和该当前订单的始发地与目的地之间的距离相关；以及基于该抢单概率，向该服务请求者发送该当前订单。
这样，能够减少该服务请求者抢单概率较低的订单的发送，即能够减少针对该服务请求者来说无价值或低价值的订单的发送，从而保证针对该服务请求者来说高价值的订单的快速地、精准地发送。
如上文已经详细描述的，该服务请求者既可以包含传统意义驾驶车辆、船、飞行器的服务提供者，也可以包含无人驾驶时用于载客/载物的交通工具。
根据本实施例，首先可以分别获取当前订单的始发地与服务请求者的位置，然后计算该当前订单的始发地与该服务请求者的位置之间的距离。其中，该当前订单的始发地可以从上述订单信息中来获取；该服务请求者的位置可以经由该服务请求者的服务请求者设备中的定位系统定位信息和/或基站定位信息来确定。另外，对于该当前订单的始发地与该服务请求者的位置之间的距离，既可以是它们之间的直线距离，也可以是当它们被置于导航系统中时参考路线信息、路况信息和路政信息而计算得到的车辆实际行驶距离。接下来，获取该服务请求者对于历史订单的抢单概率，其中该历史订单的始发地与该服务请求者之间的距离和该当前订单的始发地与该服务请求者之间的距离相关。例如，当存在大量历史订单使得每个历史订单的始发地与该服务请求者之间的距离和该当前订单的始发地与该服务请求者之间的距离都相关时，既可以分别获取该服务请求者对于每个历史订单的抢单概率，也可以获取该服务请求者对于这些历史订单整体的抢单概率。
根据本实施例，该历史订单的始发地与该服务请求者之间的距离和该当前订单的始发地与该服务请求者之间的距离的相关性可以体现如下：
该历史订单的始发地与该服务请求者之间的距离等于该当前订单的始发地与该服务请求者之间的距离。
在这一实施例中，因为该历史订单的始发地与该服务请求者之间的距离等于该当前订单的始发地与该服务请求者之间的距离，所以该服务请求者对于该历史订单的抢单概率在很大程度上等于该服务请求者对于该当前订单的抢单概率。也就是说，如果该服务请求者对于该历史订单的抢单概率较低，则该服务请求者对于该当前订单的抢单概率在很大程度上也较低。这样，该当前订单对于该服务请求者来说将可能是无价值或低价值的，因此该当前订单的发送将可能影响针对该服务请求者来说高价值的订单的发送。因此，在这一实施例中，通过减少该服务请求者抢单概率较低的当前订单的发送，能够保证针对该服务请求者来说高价值的订单的快速地、精准地发送。
(2)该历史订单的始发地与该服务请求者之间的距离和该当前订单的始发地与该服务请求者之间的距离属于相同的距离区间，其中该距离区间按照各个历史订单的始发地与服务请求者之间的距离而预先分配。例如，0-100米是第一距离区间，通过P1来表示；100-200米是第二距离区间，通过P2来表示；200-300米是第三距离区间，通过P3来表示；以此类推。
在这一实施例中，因为该历史订单的始发地与该服务请求者之间的距离和该当前订单的始发地与该服务请求者之间的距离属于相同的距离区间，所以该服务请求者对于该历史订单的抢单概率在很大程度上近似于该服务请求者对于该当前订单的抢单概率。也就是说，如果该服务请求者对于该历史订单抢单概率较低，则该服务请求者对于该当前订单的抢单概率在很大程度上也较低。这样，该当前订单对于该服务请求者来说将可能是无价值或低价值的，因此该当订单的发送将可能影响针对该服务请求者来说高价值的订单的发送。因此，在这 一实施例中，通过减少该服务请求者抢单概率较低的当前订单的发送，能够保证针对该服务请求者来说高价值的订单的快速地、精准地发送。
输入信息预处理
在基于服务提供者与服务请求者的信息，对服务信息进行派发之前，需要对输入到系统的信息进行预先的处理。
一般地，在服务派发系统中，由于来自服务提供者、服务请求者以及外界信息在输入方式上存在不同，例如文本输入、语音输入、图像输入等不同的信息输入方式，输入信息的格式上很难统一；或者，由于信息输入的不确定性或不规范性，造成输入信息难以识别，如手写输入中难以识别的笔迹，语音输入中难以识别的方言、腔调，图像输入中难以识别的图像元素。因此，在服务派发系统中，为了处理模块后续对于输入信息的计算、判断、匹配等处理步骤的实施，一般地，需要对输入的信息进行预处理。预处理的方式一般包括特征信息的识别与提取，以及信息格式的转化等。
对于输入信息的预处理，可以包括对输入的文本信息的预处理。文本信息，一般地，包括服务请求者输入的地址信息，身份信息，订单需求信息等。服务派发系统接收到服务请求者输入的信息后，需要识别输入信息，提取特征信息，对信息格式进行转化。例如，在许多依赖于地理位置信息的场景中，服务提供者和服务请求者对于同一地点的表达倾向不同，理解也不同，这就导致服务请求者所表述或指明的地点难以被服务提供者所理解，导致服务请求难以被服务提供者接受或执行。
服务派发系统，能够将服务请求者和/或服务提供者输入的一个地址信息(又称地理兴趣点，即point of interest，简写为poi)改写或转换为执行服务内容的服务提供者(俗称接单人)所能够理解的地理兴趣点，以便缩短对接单人的地点播报，清晰化非知名地点(比如不知名小区)，同时提供简短易懂的地点描述。处理模块对输入的地址信息进行预处理。
具体地，对于输入的文本格式的地理兴趣点信息(例如，纽约市公共图书馆Schwarzman大楼)，服务派发系统可以将该信息转换为包含有一个起到标识作用的展示名称(如，纽约市公共图书馆主馆Stephen A.Schwarzman大楼)和一个已存在于云端服务器的地址条目库中的详细地址(如，纽约市42街/第五大道)的文本，从而作为所述地址改写模块的输入信息。
对输入信息的预处理还可以包括对输入的语音信息的预处理。语音信息，一般地，包括服务请求者通过智能设备输入的声音，声音内容可以包括对服务的需求，服务请求者自身的身份信息等。处理模块接收到服务请求者输入的语音信息后，需要识别输入的语音信息，提取特征信息，对信息格式进行转化。例如，在食物配送服务场景中，外卖用户(服务请求者)通过语音方式呼叫餐厅外卖配送人员(服务提供者)，用以告知餐厅外卖配送人员送餐时间，送餐地点，菜品信息。但是，一般地，语音信息存在者声调，口音，断句，结巴等可能出现的问题，很有可能导致外卖配送人员(服务提供者)听不清，听不懂或者听错。处理模块，能够将外卖用户(服务请求者)的语音信息，进行识别，并提取出特征信息。
具体地，语音信息，可以通过语音识别设备，对语音信息中包含的信息进行分离、识别，以达到将输入的语音输出为对应的文本信息，并对输出的文本信息进行文本信息的预处理。语音信息转换为文本信息的过程，通常称为语音到文本(STT)。STT的步骤可以大略分为语音的预处理与音频的文字识别。语音识别设备，可以采用有限傅立叶变换、小波傅立叶变换、离散傅立叶变换、卷积、小波分析、滤波、降噪等技术人员所熟知的处理步骤，对音频进行预处理。在预处理后，处理模块对音频信息采取包括但不限于模式识别、模式对比、发音词典查找、解码、以输出文本信息。最后，在输出文本信息后，处理模块，还可以结合语言模型、语法规则以及语音识别的历史数据与现有模式，将文本信息整理为符合语法规则的文本内容。
处理模块对于输入信息的预处理还可以包括对输入的图像信息 的预处理。图像信息，一般地，包括服务请求者通过智能设备拍摄或输入的照片、视频、二维码等。图像信息可以包括服务需求信息、服务请求者个人信息、服务请求者自身的地址信息、身份信息等。处理模块接收到服务请求者输入的图像信息后，需要识别输入的图像信息，提取特征信息，对信息格式进行转化。例如，在上门洗衣服务场景中，洗衣用户(服务请求者)通过拍摄衣物照片告知洗衣房取件人员(服务提供者)衣物的具体状况。但是，一般地，服务请求者或服务提供者所输入的图像信息可能存在清晰度差、色差、畸变、抖动、模糊等问题，很有可能导致另外一方难以获取有用信息。处理模块，能够将洗衣房取件人员的图片信息，进行识别，并提取出特征信息。
以交通服务为例，通常的服务订单中包含多个地址信息，例如，乘客当前位置、司机当前位置、乘客目的地等。然而，司机和乘客对于同一地点的表达倾向不同、理解也不同。在一些情况下，司机对于商圈、路名、地标性建筑比较熟悉，而乘客则倾向于更精细、准确的表述，比如小区名、大厦名、商户名等等，这就导致乘客表述的地点司机难以理解，不愿意抢不知道地点在哪里的订单，直接影响司机的抢单意愿。
在本发明的一个实施例中，处理模块，能够将服务请求者(比如需要交通服务的个人)输入的一个地址信息(又称地理兴趣点，即point of interest，简写为poi)改写或转换为服务提供者(比如司机)理解偏好的地理兴趣点，以便缩短对服务提供者的地点播报，清晰化非知名地点(比如不知名小区)，同时提供简短易懂的地点描述。
处理模块对输入的地址信息进行预处理。所述预处理过程，用于将输入的单条地理兴趣点文本转换为展示名称和详细地址两条文本，从而作为所述地址改写模块的输入信息。
图10为预处理地址信息的流程图。
在步骤S500，接收包含展示名称和/或详细地址的信息；
在步骤S510，判断所述信息是否为展示名称和详细地址混编的单条文本；若是，转下一步骤S520；若否，转步骤S630，展示包含 展示名称和详细地址的信息；
在步骤S520，将信息切割为展示名称和详细地址；
在步骤S530，展示包含展示名称和详细地址的信息。
其中，所述预处理模块的步骤S510中，如果该地理兴趣点的展示名称与详细地址相同，或者展示名称不为空，而详细地址为空，则属于该地理兴趣点文本为展示名称和详细地址混编的单条文本。
进一步优选的，所述预处理模块的步骤S520是按照区字将该地理兴趣点的展示名称或详细地址切割为双条记录包括展示名称和详细地址；
作为一种特殊情况，如果在切割完后，区字前面是小、一、二、三、四、五、六、七、八、九、十、东、南、西或北字，放弃切割；
切割完后，如果该地理兴趣点的展示名称长度小于3个汉字的放弃切割。
优选的，处理模块还对输入的包含展示名称和详细地址的poi改写为接单人理解偏好的poi。
处理模块还可以改写服务请求者发出的服务请求信息中包含的地址信息。图11为改写地址信息的流程图。
在步骤S600，接收包含展示名称和详细地址的信息poi；
在步骤S605，判断该poi是否含有一地标性建筑。如果是，清空address，转步骤S620；如果否，则转入下一步骤S610；
在步骤S610，判断poi中displayname(展示名称)是否包含站、村、桥、地铁、立交或机场等通用地点名称关键字；如果是，则清空address，转步骤S620；如果否，则转入下一步骤S615；
在步骤S615，判断poi中displayname(展示名称)是否包含路、街或道等关键字，或者判断displayname(展示名称)长度是否大于8个汉字；如果是，不予处理，转步骤S620；如果否，则转入下一步骤S630(displayname包含路、街或道等字时，说明它自己足够表明自己的位置，比如displayname：“中关村大街11号”；displayname过长的情况下说明它自己足够表明自己的位置，比如displayname：“中 关村e世界c座，中钢国际对面”)；
在步骤S620，清空详细地址后，进入步骤S625；
在步骤S630，反解析经纬度，对于poi没有经纬度的(比如，乘客手动输入“中关村”，没有带经纬度信息)，通过该poi的displayname或address从地址经纬度集合中解析出该poi对应的经纬度(poi有经纬度的则忽略此步骤)，计算起点终点距离；
在步骤S635，判断起点poi和终点poi的经纬度距离是否大于限制距离(例如，限制距离为50000米～60000米)，如果是，不予处理，转步骤S625；如果否，则转入下一步骤S640(本步骤是为防止经纬度出错的预防步骤，经纬度信息准确的，也可忽略此步骤)；
在步骤S640，解析商圈信息或地址信息，通过该poi的经纬度从商圈经纬度集合中解析出该poi对应的商圈信息，或者区和道路等地址信息；判断是否获取商圈信息，如果是，将address设置为商圈，address＝商圈，转步骤S645；如果否，则转入下一步骤S650；
在步骤S645，将详细地址设置为商圈，并进入步骤S625，输出展示名称和详细地址；
在步骤S650，判断是否获取区和道路信息，如果是，将address设置为区+道路信息，比如“朝阳区延静里中街”，转步骤S655；如果否，则转步骤S625，输出展示名称和详细地址；
在步骤S655之后，转步骤S625；
在步骤S625，输出包含展示名称和详细地址的poi。
所述步骤S605中判断poi是否含有地标性建筑的具体方法为：
设置地标性建筑信息集合，所述地标性建筑信息集合中的各条记录包括：编号、地址信息、点击数，所述地标性建筑信息集合用于收集乘客对于各个地址信息的点击数；
设定地标性建筑点击数阈值Tclick(该阈值可以根据各个城市的实际情况进行设定)；
通过该poi的displayname或address从地标性建筑信息集合中查找该poi的点击数，如果该点击数大于Tclick，则该poi为地标性建 筑。
所述地址经纬度集合中的各条记录包括：编号，地址，经纬度。
所述商圈经纬度集合中的各条记录包括：编号，商圈名称和/或区道路信息，经纬度。
优选的，处理模块用于进一步缩短poi中的address，删除不妨碍接单人理解的前缀和后缀，比如城市名前缀，街牌号后缀等等。
图12为缩写地址信息的流程图。
在步骤S700，输入包含展示名称和详细地址的poi；
在步骤S710，判断poi中的详细地址的前缀是否为城市，如果是，转下一步骤；如果否，转步骤S730；
在步骤S720，切除城市前缀，比如北京市，福建省漳州市，因为现有的订单分配通常只限于单个城市，所以城市信息是冗余的；
在步骤S730，判断poi中的详细地址的末尾是否为门牌号，如果是，转下一步骤；如果否，则转步骤S750；
在步骤S740，切掉末尾门牌号(比如“中关村大街11号”，处理后变为“中关村大街”；对于接单人而言，听单的时候主要是通过目的地点的大概区域判断是否有接单需求，详细的地点可以之后再了解)；
在步骤S750，输出包含展示名称和详细地址的poi。
应用场景
以终点Address＝“北京市海淀区中关村大街15号新中关购物中心地下1层”，displayname＝“金逸国际影城”为例，这个poi非常长，需要花费10秒左右的时间才能播放完毕，一方面浪费了司机的时间，另一方面冗余的信息其实对司机来说并没有用。
经过地址改写模块和地址缩写模块的判定，这个poi需要被改写和缩写。地址改写模块提取出最有用的商圈信息“中关村”，地址缩写模块识别这里的无意义信息“北京市”、“15号”、“地下1层”。最后，将这个终点poi改写为Address＝“中关村”，displayname＝“金逸国际影城”。
应用场景
以终点Address＝空，displayname＝金泰富地大厦为例。这个终点非常短，并且没有地址信息，只有一座并非地标、很少有司机知道的建筑名称。这就导致司机无法判断终点方位，从而影响了抢单意愿。
经过地址改写模块的判定，这个终点poi的详细地址需要被改写。通过改写模块会查到这个建筑位于海淀区安宁庄路附近，因此这个终点poi会被改写为Address＝安宁庄路，displayname＝金泰富地大厦。
以上是对处理模块在派发订单前根据历史数据、服务请求者/服务提供者输入的信息进行预处理与计算的描述。在根据历史数据，挖掘到服务请求者/服务提供者的偏好、服务范围、常驻点、营业时间、忙碌时段等信息后，服务派发系统可以利用这些信息，结合具体的判据，为不同的服务请求信息，派发给不同的服务提供者，从而实现高效率，高针对性的服务派发。
服务订单的拍卖
现有的服务系统存在如下两个问题：1、当前服务提供者获得订单的模式为抢单模式，智能设备网络条件较佳、抢单动作较快者优先获得订单，但是先抢单到订单的服务提供者可能为距离服务请求者较远或服务较差的服务提供者，反而优质服务提供者仅因为抢单速度慢而无法接单，导致服务请求者无法获得更优质服务提供者的服务。例如，到家医疗服务场景中，病人下单后，可能订单被一些抢单速度快但医疗效果或者评价并不好的医疗机构或个人抢单；2、在订单争抢过程中，滋生了利用网络接入速度、硬件性能等的作弊工具，破坏了平台竞争的公平性。
本实施例，可以看做一种基于拍卖模式获得优质服务提供者的派单系统。该系统包括两个步骤，排序步骤和拍卖步骤。
所述排序步骤用于根据订单特征指标、服务提供者信息等因素筛选出某一订单的服务提供者集合并进行排序；
所述拍卖步骤，用于在所述判断模块筛选和排序出的服务提供者集合中，对所述订单进行拍卖抢单。
优选的，设所述订单已经由处理模块根据其特征指标以及n个服 务提供者D1、D2、…、Dn的信息等因素计算，确定了一个服务提供者的排序：D1>D2…>Dn，Dx为其中某个服务提供者。
处理模块为所属订单指定了一个拍卖时间阈值T，为每个服务提供者Dx指定了一个拍卖时间tx，tx可以随着服务提供者Dx的不同而有所差异。
参照图13，所述拍卖抢单的步骤包括：
步骤S800：司机Dx发起抢单；
步骤S810：获取司机Dx在整个序列D1>D2…>Dn中的排序；
步骤S820：计算拍卖时间tx；
Figure PCTCN2015086075-appb-000001
步骤S830，判断Tx是否为0，当Tx为0时，跳至步骤S840，通知Dx抢单成功，并通知剩余司机拍卖失败，拍卖结束，当Tx不为0时，执行S850；
步骤S850：通知排名低于Dx的司机Dx+1～Dn拍卖失败和失败原因，同时通知排名高于Dx的司机D1～Dx-1继续参与拍卖，且拍卖时间为Tx；
当有排名高于Dx的司机参与拍卖时，跳转至步骤S800，循环。
上述技术方案，可以达到如下技术效果：
1、将原有派单系统中司机手快先得的模式升级为通过轻量级拍卖获得最优解的模式，使订单分配给最优质的司机，缩短乘客等待时间、并使乘客获得最优服务；
2、可以使司机间的竞争更加公平，使司机的努力方向更加符合平台的需求、提高司机的抢单积极性，从而提升全平台的服务质量。
应用场景
某北京乘客下班期间使用打车软件预定从西直门到三里屯的出租车，获取到的乘客信息为：
乘客 叫单时间 当前所在地点 目的地
李某 2014/2/20 18:00 西直门 三里屯
收集到的若干司机信息如下：
司机编号 服务质量(以100为满分) 司机当前位置 司机等级 其他信息
1 60 大钟寺 B ……
2 50 西直门 B ……
3 40 知春路 D ……
4 80 知春里 C ……
5 70 北土城 A ……
6 80 北土城 B ……
7 70 安贞桥 C ……
8 80 海淀黄庄 A ……
9 40 动物园 D ……
10 65 南锣鼓巷 C ……
计算司机当前位置与乘客的距离，转化为如下表格：
司机编号 服务质量(以100为满分) 司机与乘客的距离 司机等级 其他信息
1 60 1.3km B ……
2 50 500m B ……
3 40 2km D ……
4 80 2.5km C ……
5 70 3km A ……
6 80 2.5km B ……
7 70 3.3km C ……
8 80 3km A ……
9 40 3km D ……
10 65 2.5km C ……
然后以司乘距离、司机服务质量和司机等级中的任意一项为权重，其余两项为副选项对司机进行排序；
权重是一个相对的概念，是针对某一指标而言。某一指标的权重是指该指标在整体评价中的相对重要程度。
如果司乘距离的权重系数为1，则司机服务质量和司机等级的权重系数为0；如果司机服务质量的权重系数为1，则司乘距离和司机等级的权重系数为0；同理如果司机等级的权重系数为1，则机服务质量和司乘距离的权重系数为0。司机服务质量可以直接由乘客对司机的历史评价得出(乘客每次打完车都可以给司机评分)；
司机等级是由司机接单的数量评定的；例如：司机接单量小于100单的等级为D，大于等于100单小于200单的等级为C，大于等于200单且小于300单的等级为B，大于等于300单的等级为A；
在第一个实施例中，是以司机与乘客的距离为优先项，服务质量为第二项，司机等级为第三项，进行排序，得到如下表格：
司机编号 司机与乘客的距离 服务质量(以100为满分) 司机等级
2 700m 60 B
1 1.3km 50 B
3 2km 40 D
4 2.5km 80 C
10 2.5km 65 C
8 3km 80 A
5 3km 70 A
9 3km 40 D
7 3.3km 70 C
6 3.5km 80 B
在第二个实施例中，是以司机与乘客的距离为优先项，司机等级 为第二项，服务质量为第三项，进行排序，得到如下表格：
司机编号 司机与乘客的距离 司机等级 服务质量(以100为满分)
2 700m B 60
1 1.3km B 50
3 2km D 40
4 2.5km C 80
10 2.5km C 65
8 3km A 80
5 3km A 70
9 3km D 40
7 3.3km C 70
6 3.5km B 80
在第三个实施例中，是以司机服务质量为优先项，司机与乘客的距离为第二项，司机等级为第三项，进行排序，得到如下表格：
司机编号 服务质量(以100为满分) 司机与乘客的距离 司机等级
4 80 2.5km C
8 80 3km A
6 80 3.5km B
5 70 3km A
7 70 3.3km C
10 65 2.5km C
1 60 1.3km B
2 50 700m B
3 40 2km D
9 40 3km D
在第四个实施例中，是以司机服务质量为优先项，司机等级为第二项，司机与乘客的距离为第三项，进行排序，得到如下表格：
司机编号 服务质量(以100为满分) 司机等级 司机与乘客的距离
8 80 A 3km
6 80 B 3.5km
4 80 C 2.5km
5 70 A 3km
7 70 C 3.3km
10 65 C 2.5km
1 60 B 1.3km
2 50 B 700m
3 40 D 2km
9 40 D 3km
在其他实施例中，也可以以司机等级为优先项，服务质量和司机等级中的一项作为第二项，剩余一项作为第三项进行排序，具体的排序方式在此不做具体限定。
从乘客的角度看，离乘客越近的司机越优质，如一个离乘客100米的司机接单，通常优于离乘客1公里的司机。
还需要考虑司机服务等级，司机历史服务越好的，对乘客而言越有吸引力，同理司机信用越好的，越能促进订单成交。
同时还需要考虑行驶路况，两个司机一个来自于拥堵区域，一个来自于畅通区域，对于乘客来说优质程度也有差别。
拍卖时间与排序呈负相关；排序D1>D2>…>Dn，当司机Dn抢单时，拍卖时间增加，排序高于Dn的司机可在剩余的时间内抢单当司机D1抢单时，拍卖时间缩短，拍卖结束后将等待拍卖的司机释放 参与下一场拍卖。
例如订单发给10个司机，拍卖时间范围是[0,7]、分割为10段，top1的司机抢单后拍卖时间就是0，top2抢单后拍卖时间就是0.7s，……，top10抢单后拍卖时间为6.3s。
Figure PCTCN2015086075-appb-000002
应用场景
举例说明，在订单分配中，记录了一个订单被推送过哪些司机，在打分模块中，记录了订单—司机对的评分。假设一个订单被推送给10个司机，按照优质程度排序为司机1、司机2、…、司机10，其拍卖流程如下：
1)假设司机1第一个抢单
该司机已经是最优司机，则拍卖直接结束。告诉其他司机，该订单已结束，订单被司机1获得，其匹配程度最高，乘客收到司机1成功抢单的消息。
2)假设司机5第一个抢单
A.对司机6～司机10而言，他们即使抢单也比不过司机5，因此对于司机6～司机10而言，订单已结束，订单可能被司机5获得，乘客收到订单已被抢的消息，但是具体谁被抢待定。
B.对于司机1～司机4而言，若他们抢单，会可能赢得抢单，因此还会在一定的时间内继续有抢单的机会。
C.假设在这段时间内，司机2抢单，同理对于司机3～5，订单已结束，订单可能被司机2获得，司机1还可以继续抢单。
D.假设司机2抢单后，司机1不抢单，订单结束，乘客收到司机2成功抢单的消息。
E.假设在这段时间内，司机1抢单，则订单直接结束，乘客收到司机1成功抢单的消息。
服务订单的积分分配
处理模块，可以基于价值判断赋予服务订单合适的积分，通过将 积分纳入订单拍卖影响服务提供者抢单成功率，来刺激服务提供者承担或执行在其他情况下不容易被接受的服务订单。总的原则就是，如果某一服务请求难以被多数服务提供者所接受，则向该服务请求分配较高积分；如果某一服务请求，被多数服务提供者所接受，则分配较低积分。
为了实现上述目的，本实施例采用的技术方案是：一种基于订单价值判断促进订单成交的服务订单积分系统，包括两步：订单积分分配步骤以及异常处理步骤。其中，订单积分分配根据历史数据判断某一订单的订单价值并进行相应的积分分配；异常处理，用于在所述订单被某一服务提供者抢单后，根据服务提供者和服务请求者行为和相关坐标轨迹进行判断，确定积分发放。
所述历史数据来自订单集合，所述订单集合包括：叫单时间、服务提供者与服务请求者距离。
以交通服务为例，订单积分分配包括如下步骤：
步骤一：预设一时间周期T，在所述时间周期T内设置若干时间点T1、T2、…、Tn，所述时间点将所述时间周期分成若干时间区间0～T1、T1～T2、…、Tn-1～Tn；
步骤二：给每一段时间区间分配若干抢单概率；
步骤三：在每一段时间区间中，根据不同的抢单概率，给予不同的订单相应的得分价值，所述抢单概率来自预先计算的概率集合；
步骤四：预设一时间周期T＝180s，在所述时间周期T内设置5个时间点30s、60s、90s、120s、150s，所述时间点将所述时间周期分成6段时间区间，分别为0～30s、30s～60s、60s～90s、90s～120s、120s～150s、150s～180s；
步骤五：给每一段时间区间分配若干抢单概率；
步骤六：在每一段时间区间中，根据不同的抢单概率，给予不同的订单相应的得分价值。
以上的列举仅仅是说明性的，并不用于限制本发明，可以理解，时间周期的长度、区间划分长度、区间数量等细节都可以随着交通服 务情况的变化而调整。
优选的，所述时间区间，进一步为订单生命周期内的某一段时间区间。
优选的，所述积分分配模块，进一步为，对所述订单判断模块得出的高价值订单分配负积分，对所述订单判断模块得出的低价值订单分配正积分。
优选的，所述订单集合来自叫车乘客信息集合，所述叫车乘客信息集合至少包括：叫单时间、订单行驶距离和乘客位置信息。
所述异常处理包括如下步骤：
步骤一：司机抢单成功后，开始收集司机和乘客地理位置轨迹；
步骤二：当司机通过服务应用，确定已接到乘客时，根据当前司机与乘客地理位置，判断服务提供者是否接到乘客，如果确认服务提供者接到乘客，继续记录司机与乘客地理位置轨迹；
步骤三：当乘客通过服务应用，确认已抵达目的地或已完成支付时，根据当前乘客地理位置，判断司机是否将乘客送达目的地；如果乘客坐标位于目的地，则发放积分；如果乘客坐标不在目的地，该订单视为作弊订单，扣除司机的积分。
以交通服务为例，图14是积分操作流程图。
在步骤S900，获取订单信息，该订单信息，包含有交通服务的起点位置信息、终点位置信息、请求服务的时间，进一步地，可以包含对司机(服务提供者)一方的要求，如如所驾驶的交通工具的类别(包括车型，如小汽车、SUV、吉普；动力类型，如电动、内燃机动力；耗油量；马力；品牌等)，可以包含小费信息，还可以包含请求交通服务的原因，如前往机场乘飞机，前往医院等；
步骤S910，计算订单积分，根据订单信息，计算订单所对应的积分，一般地，积分受到多种参数的影响，如订单的起点与终点各自的位置，起点与终点之间的路程长度和/或距离，交通工具的类型，小费数量，交通服务的紧急程度或原因；
步骤S920，接收获取并执行订单服务内容的司机信息，在确认 执行服务内容后，司机的信息，如年龄、性别、犯罪记录、诚信记录、从业时间等信息，可以通过系统调取云端的记录而获得；
步骤S930，判断是否发生了异常事件，若否，进入S940，订单已完成，发放所计算的订单积分；若是，进入S950，订单未完成，不发放所计算的订单积分。这里，异常事件是指，多种原因导致的交通服务的未完成，如司机恶意诱骗乘客确认订单完成，司机伙同乘客作弊等。这些异常事件，可以通过车载定位装置与乘客智能设备中的定位模块提供的车辆位置与乘客位置所确定。
尽管本实施例以交通服务为例，说明了积分操作的流程，但是可以理解，其他服务，例如送餐服务、洗衣服务、送货服务，都可以遵循类似的积分操作，为服务分配积分，并在服务提供者执行并完成服务请求所涉及的内容后，为其发放积分，并在发现服务提供者的异常事件时，扣除其积分。
以下，将通过实施例，结合不同判断标准与相应判据，对判断流程进行说明。
顺风车订单判断
现有的交通服务分配系统在进行交通服务订单分配的时候，一般情况下是把交通服务订单推送给全部出租车，由各个出租车司机进行抢单，而没有考虑到各个出租车司机个性化的需求。比如，司机在出班和收班的时候，更倾向于接在时间上符合上下班时间和空间上在家附近的订单；而现有的交通服务订单分配方法，并没有考虑一个订单的出发时间、出发地点是否与某个或某些司机出班和收班的时间地点相关联，而是将此时间地点的订单作为普通时间地点订单一起进行订单分配，这样司机在回家的时候往往分不到或抢不到和自己顺路的订单。
顺风车，通常又称为搭便车。如果，乘客所寻求的交通服务的方向，与司机当前或在未来一段时间内所行驶的方向相同或接近，则该乘客的交通服务需求，构成一个顺风车订单。常驻点，在本文中应当理解为服务提供者在一定时间内经常停留、到达的地点，例如家庭住 址、自己或配偶的工作地点、常去的就餐地点、孩子就读的学校、常去的加油站、常去的健身馆/运动馆/游泳馆等。
本实施例能实现基于司机常驻点信息的顺风车订单识别功能，通过顺风车规则，向司机推送在诸如出班、收班、外出等前往常驻点或离开常驻点等情形下时间和距离上更贴近的订单，提高系统派发订单的针对性，提高司机接单的效率和准确性。
本实施例能实现基于司机住址信息的顺风车订单判断功能。通过分析来自司机的信息与乘客需求的交通服务信息，判断订单是否是司机的顺风车订单。其中，来自司机的信息，包括司机的当前位置，司机住址的信息。进一步地，还包括司机编号，出班位置或收班位置，出班时间或收班时间。进一步地，司机的当前位置，可以通过位于司机车辆中的智能设备中的定位模块获得，并发送至系统。进一步地，乘客需求的交通服务信息包括订单编号、起点位置、终点位置、出发时间。
本实施例中，系统通过分析来自司机的信息与乘客需求的交通服务信息，判断订单是否是司机的顺风车订单。其中，来自司机的信息，包括司机的当前位置，司机常驻点等信息。进一步地，司机的当前位置，可以通过位于司机车辆中的智能设备中的定位模块获得，并发送至系统；司机常驻点的信息，可以通过对司机所在位置、运行轨迹的历史数据而得到。
顺风车判定
图15为处理模块进行出班顺风车判定的流程图。
在本实施例中，处理模块130具体的功能为针对某个特定的司机，判定当前交通服务请求者所寻求的交通服务是否构成顺风车订单。
处理模块130，根据来自信息收集模块110的信息，具体地，来自乘客的交通服务的需求信息与来自司机的交通服务的提供信息，并结合系统所存储的司机历史位置、历史轨迹或输入或系统计算得到的常驻点信息，来判断乘客的交通服务需求是否属于该司机的顺风车订单。
所述各交通服务订单信息包括：订单编号、起点位置(如起点经纬度)、终点位置(如终点的经纬度)、出发时间。以上信息并非订单信息的全部，其他一些与订单有关的信息，例如，寻求交通服务的原因，对交通服务提供者的偏好或要求等，例如，交通服务提供者的年龄、性别、驾龄、交通工具的类型、性能等，也可能被包括在内。
所述各出租车信息包括：司机编号，当前位置(如当前经纬度)，出班位置(如出班经纬度)或收班位置(如收班经纬度)，出班时间或收班时间。以上信息并非订单信息的全部，其他一些与订单有关的信息，例如，对交通服务请求者的偏好或要求等，例如，对所携带行李的要求或限制，也可能被包括在内。
根据出班顺风车判定规则计算出班顺风车订单：
所述各出租车信息包括：司机编号，当前位置，出班位置，出班时间；
在步骤S1000，取叫车订单集合中任一叫车订单Q，取出租车信息集合中任一出租车信息C；
在步骤S1010，获取订单Q出发时间T，订单距离D，出租车信息C中出租车出发时间Tout，出租车出班时间阈值TYout，订单Q的起点与出租车的出班位置距离为Dout，出租车出班距离阈值DYout，出班距离倍数阈值Kout；
在步骤S1020，若T减去Tout的绝对值小于或等于TYout(|T-Tout|<＝TYout)，则认为订单Q满足条件1，进入S1030；若T减去Tout的绝对值大于TYout(|T-Tout|<＝TYout)，则认为订单Q不满足条件1，进入S250，判定该订单不是该司机的出班顺风车订单；
在步骤S1030，若Dout小于或等于DYout(Dout<＝DYout)，或者订单距离D大于或等于Kout乘以DYout(D>＝Kout*DYout)，则认为订单Q满足条件2，进入S1040，判定该订单为该司机的出班顺风车订单；否则，进入S1050，判定该订单不是该司机的出班顺风车订单；
出班时间阈值TYout取值[0.5,1.5]小时，即0.5小时至1.5小时， TYout取值小则获选的匹配订单数量较少或没有，TYout取值大则司机出班时间提前或拖延太长。
出班距离阈值DYout取值[3000,5000]米，即3000米至5000米，取值小则获选的匹配订单数量较少或没有，取值大则司机空驶的路程可能就较多。
出班距离倍数阈值Kout取值[2,30]倍，即2倍至30倍，在出班距离阈值DYout取值固定情况下，Kout取值小则获选的匹配订单数量较多，Kout取值大则获选的匹配订单数量较少或没有。
图16为处理模块进行顺风车判定的另一个流程图，在这个流程中，一个订单被判断，决定是否构成收班顺风车。所述出租车信息包括：司机编号，当前位置，收班位置，收班时间；
在步骤S1100，取叫车订单集合中任一叫车订单Q，取出租车信息集合中任一出租车信息C；
在步骤S1110，获取订单Q出发时间T，订单距离D，出租车信息C中出租车收班时间Tin，出租车收班时间阈值TYin，订单Q的起点与出租车的当前位置距离为Din1，出租车的收班位置与当前位置距离为Din2，订单Q的终点与出租车C的收班位置距离为Din3，订单Q的订单距离为D，第一收班距离阈值DYin1，第二收班距离阈值DYin2，收班距离倍数阈值Kin；在步骤S1120，若T减去Tin的绝对值小于或等于TYin(|T-Tin|<＝TYin)，则认为订单Q满足条件1，进入步骤S1130；否则，进入步骤S1160，判断该订单不是该司机的收班顺风车订单；
在步骤S1130，若Din1减去Din2大于等于DYin1(Din1-Din2>＝DYin1)，则认为订单Q满足条件2，进入步骤S1140；否则，进入步骤S1160，判断该订单不是该司机的收班顺风车订单；
在步骤S1140，若Din3不大于DYin2(Din3<＝DYin2)，或者订单距离D大于或等于Kin乘以Din3(D>＝Kin*Din3)，则认为订单Q满足条件3，进入步骤S1150，判断该订单是该司机的收班顺风车订单；否则，进入步骤S1160，判断该订单不是该司机的收班顺风车订 单。
需要注意的是，步骤S1120、S1130、S1140的顺序可以颠倒变换，同时，这几个步骤不是全都必需的，可以根据需要，删除一个或者两个判断步骤，而不影响在后续步骤中对订单是否构成收班顺风车的判断。例如，可以只判断S1120中的条件是否满足，而决定订单Q是否是出租车C的收班顺风车，而不再进行S1130和/或S1140中条件的判断。
收班时间阈值TYin取值[0.5,1.5]小时，即0.5小时至1.5小时，TYin取值小，则获选的匹配订单数量较少或没有，TYin取值大，则司机收班时间提前或拖延太长。
第一收班距离阈值DYin1，可以有不同的取值范围，作为一个示例，DYin取值为[1000,5000]米，即1000米至5000米。DYin1取值小，则获选的匹配订单数量较多，DYin1取值大，则司机空驶的路程就较多。
第二收班距离阈值DYin2可以有不同的取值范围，作为一个示例，DYin2取值[3000,5000]米，即3000米至5000米。DYin2取值小，则获选的匹配订单数量较少，DYin2取值大，则司机空驶的路程就较多。
收班距离倍数阈值Kin取值[2,10]倍，即2倍至10倍，Kin取值小则获选的匹配订单数量较多，Kin取值大则获选的匹配订单数量较少或没有。
以上对出班时间阈值、出班距离阈值、出班距离倍数阈值、收班时间阈值、第一收班距离阈值、第二收班距离阈值、收班距离倍数阈值的描述仅仅是示例性的，而非出于限制本发明中相应概念的目的，可以理解，任何适合的阈值范围都可能被本领域的技术人员所构想，这些所构想的阈值范围都不脱离本发明的所寻求保护的范围与精神。
在为特定司机计算并判断出顺风车订单后，处理模块将顺风车订单的判断结果输出至输出模块，以供其推送给该指定司机，或者将该判断结果提供给一个第三方，以供司机查询或获取。
应用场景
以北京为例，司机出班时间是上午6:00，司机家的地址为中关村大街10号。
在叫车软件的服务器端或呼叫中心的服务器端，都存储有大量的从乘客那那里收集的叫车订单。作为一种示例，从乘客收集到的叫车订单格式如下：
订单编号 乘客手机号 出发地 出发时间 出发地经纬度信息
140002 13300000001 中关村大街10号 2014/2/2 06:00 xxxxxx
140012 13300000002 中关村大街20号 2014/2/2 18:00 xxxxxx
正常情况下，每当有一个乘客发出用车请求，就会有如上的一个订单信息发送到服务器上。
每个出租车在行驶过程中，通过出租车司机的智能设备，每隔一段时间(如10秒钟)向服务器上报一次当前出租车所在位置经纬度，信息格式如下：
司机编号 上报时间 当前所在地点 经纬度信息
12345 2014/2/2 06:00 中关村大街10号 xxxxxx
基于当前的订单信息和司机信息，服务器会对订单和司机进行匹配，根据出班顺风车规则，订单140002对于司机12345，属于出班顺风车，播送该订单时相应的推送信息设为“顺风车订单”，提示司机。
应用场景
以北京为例，司机收班时间是下午6:00，司机家的地址上地三街10号。
从乘客那里收集到的实时叫车订单格式如下：
订单编号 乘客手机号 出发地 出发时间 目的地
140002 13300000001 中关村大街10号 2014/2/20 06:00 知春路29号
140012 13300000002 中关村大街20号 2014/2/20 18:00 上地三街10号
每个出租车在行驶过程中，通过出租车司机的智能设备，每隔10秒钟向服务器上报一次当前出租车所在位置经纬度，信息格式如下：
司机编号 上报时间 当前所在地点 经纬度信息
12345 2014/2/20 18:00 中关村大街20号 xxxxxx
基于当前的订单信息和司机信息，服务器会对订单和司机进行匹配。
根据收班顺风车规则，订单140012对于司机12345，属于收班顺风车；播送该订单时相应的推送信息设为“顺风车订单”，提示司机。
应用场景
以北京司机王师傅为例。王师傅家住北京西北旺，每天上午9点出班，晚上8点收班。每天出班后会习惯性的打开软件，开始一天的工作。于是，每天王师傅的轨迹为“家—地点1—地点2……地点n—家”。
司机所持有的智能设备，可以获取包括司机身份识别信息、司机实时位置等在内的信息，并定期或者不定期地向外发送。作为一个示例，从司机收集到的一段时间的8条信息集合格式如下：
Figure PCTCN2015086075-appb-000003
在某些情况下，每隔一段时间，司机的轨迹就会由如上的一个信息发送到远程的服务器上。
服务器通过获取一个司机的历史轨迹后，可以按照Dbscan聚类算法计算出司机大致的常驻点。比如下表，表示司机每天晚上8点左右时，会回到如下的经纬度附近：
编号 司机手机号 时间 经度 纬度
29912132 13300000001 20:00:00 116.236723 39.543692
该地点可能就是司机的家庭住址或者配偶的工作地点。
需要注意的是，计算司机的常驻点所利用的算法并不局限于Dbscan聚类算法，可以理解，用于计算司机常驻点的算法，还可以包括其他，如划分方法(Partitioning method)，如K-means、K-medoids、CLARA(Clustering LARge Application)、CLARANS(Clustering Large Application based upon RANdomized Search)、FCM等；层次法(hierarchical method)，如BIRCH(Balanced Iterative Reducing And Clustering using Hierarchies)、正二进制(binary-positive)方法、连续数据的粗聚类算法(Rough Clustering Of Sequential Data，简称RCOSD)；基于密度的方法，如OPTICS(Ordering Points To Identify The Clustering Structure)；基于网络的方法，如STING(STatistical INformation Grid)、CLIQUE(Clustering In QUEst)、Wave-Cluster；基于模型的方法，如Cobweb、CLASSIT等。
这样，当该司机运营到19点附近时间点，就可以给他推荐去往(116.236723，39.543692)或周围若干公里方向的订单，司机对这类订单的抢单意愿会强于其他订单。
为司机合适的订单后，可以在播送该订单时，增加标签“顺风车订单”，从而提醒其这是去往司机交接地的订单。
应用场景
以北京司机王师傅为例。王师傅每天上午9点出班，晚上8点收班。每天出班后会习惯性的开启智能手机的定位功能，开始一天的工作。于是，每天王师傅的轨迹为“家—地点1—地点2—…—地点n—家”。
在叫车软件的服务器端或呼叫中心的服务器端，都存储某一司机大量上报的轨迹信息。在本场景中，从司机那里收集到的一段时间的信息集合格式如下：
Figure PCTCN2015086075-appb-000004
Figure PCTCN2015086075-appb-000005
服务器上通过获取一个司机的历史轨迹后，可以按照Dbscan聚类算法计算出司机大致的常驻点。将停留时间大于25min的经纬度保留，经过几何平均数的算法得出，司机中午吃饭的地点在如下的经纬度附近：
编号 司机手机号 时间 经度 纬度
29912132 13300000001 12:28:12 130.236723 55.543692
这样，当该司机运营到12点附近的时间，就可以给他推荐去往(130.236723，55.543692)方向的订单，司机对这类订单的抢单意愿会强于其他订单。
为司机合适的订单后，可以在播送该订单时，增加标签“顺风车订单”，从而提醒其这是去往司机交接地的订单。
司机活跃度判断
司机使用打车软件时的用户体验是与乘客是不一样的，司机的用户体验主要就是订单多少，能否抢到订单或者收到多少订单。对于体验差的司机，往往在线抢不到订单，甚至觉得没什么实际收益。因此，会因为体验差而流失部分司机。因此，判断司机的在线活跃状况，并对非活跃司机进行针对性策略运营，从而挽留住这些之前体验较差的司机，对于提升整个系统的运力是十分必要的。
本实施例的一种实现方式是通过对运营司机的筛选，找出易流失的司机，采用优质订单来召回易流失司机。为了解决上述技术问题，本实施例所采用的技术方案是：根据各司机一段时间内的在线情况和/或抢单情况，来判定各司机是否为非活跃司机。
在本实施例中，处理模块具体的功能为针对某个特定的司机，基 于该司机的一段时间内的在线情况和/或抢单情况，来判定该司机是否为非活跃司机。优选的，所述抢单情况接收自信息接收模块，叫车订单历史信息集合，所述叫车订单历史信息集合中的各订单信息包括：订单编号、出发地、目的地、出发时间、接单司机编号；
所述在线情况来自出租车信息集合，所述出租车信息集合中的各出租车信息包括：司机编号、上报时间、出租车位置；
所述非活跃司机集合中的各非活跃司机信息至少包括司机编号信息。
图17a是处理模块判断司机活跃度流程图。
进一步优选的，所述非活跃司机判定模块包括如下步骤，如图17a所示：
在步骤S1200，设任一司机A，遍历叫车订单历史信息集合和/或出租车信息集合，查询是否有司机A的记录；
在步骤S1210，如果司机A在出租车信息集合中最近T1时间段内都有记录，并且司机A在叫车订单历史信息集合中最近T1时间段内没有记录，则司机A为非活跃司机，进入步骤S1250，并将司机A记录到非活跃司机集合；否则进入步骤S1220；
所述T1时间段可以设置不同的数值，例如，5天、7天、10天、15天、30天、90天等，还可以是其他时间长度。时间段设置越长，随机因素对于司机A活跃程度判断的影响就越小，但信息量也会相应增大。
在步骤S1220，如果司机A在出租车信息集合中最近T2时间段内没有记录，并且司机A在出租车信息集合中最近T3时间段内有记录，则司机A为非活跃司机，进入步骤S1250，并将司机A记录到非活跃司机集合，否则进入步骤S1230；
作为一个示例，所述T2时间段为1～5天，所述T3时间段内为15～45天。可以理解，T2与T3还可设置为其他不同的数值，例如，5天、7天、10天、15天、30天、90天等，还可以是其他时间长度。与T1类似地，时间段设置越长，随机因素对于司机A活跃程度判断 的影响就越小，但信息量也会相应增大。
在步骤S1230，判断叫车订单历史信息集合在最近T4时间段内全体司机平均接单记录是否不大于全体司机平均接单数量的1/10-1/2，若是，则转步骤S1250，判断司机A为非活跃司机；若否，则转步骤S1240，司机A为活跃司机，所述T4时间段为1～5天；
在步骤S1260，将司机A记录到非活跃司机集合。
进一步优选的，所述叫车订单信息来自叫车订单实时信息集合，所述叫车订单实时信息集合中的各订单信息至少包括：订单编号、出发地、目的地、出发时间。
图17b是结合了非活跃司机判定的订单派发流程图。
进一步优选的，所述运营策略实现模块遍历叫车订单实时信息集合中各叫车订单信息并执行以下步骤：
步骤S1204，取叫车订单实时信息集合中的一个叫车订单信息及一个在线司机A信息；
步骤S1214，判断该叫车订单信息是否为优质订单，如果否，进入步骤S1234，向该司机播放该叫车订单信息，返回步骤S1204，取下一个在线司机信息；如果是，转下一步骤S1224；
步骤S1224，判断该订单是否被锁定，如果是，转步骤S1264；如果否，向该司机播放该叫车订单信息，转步骤S1234；
步骤S1264，判断锁定时间是否到期，锁定时间到期，向该司机播放该叫车订单信息，返回步骤S1234，向该司机播放该叫车订单信息；锁定时间未到，返回步骤S1204，取下一个在线司机信息；
步骤S1234，向司机A播放该订单信息，并跳转S1244；
步骤S1244，判断司机A是否非活跃司机，如果否，返回步骤S1204，取下一个在线司机信息；如果是，锁定该订单，并设定锁定时间，返回步骤S1204，取下一个在线司机信息。
进一步优选的，根据所述叫车订单实时信息集合中的各订单的出发地、目的地和/或出发时间信息筛选出优质叫车订单。
进一步优选的，所述筛选出优质叫车订单的步骤为：
根据订单目的地直接判断，如果订单目的地是去往机场或者火车站，则该订单属于优质订单；
或者根据订单出发地和目的地，计算出该订单的订单距离进行判断，订单距离大于X公里的订单就属于优质订单，X取值为3-10公里之间，其他的取值也是可以的，例如，20km，40km，50km等；
或者根据所述叫车订单实时信息集合中的各订单的出发地、目的地和/或出发时间信息计算各订单被抢单的可能性，预先设定抢单可能性阈值，如果一个订单被抢单的可能性大于抢单可能性阈值，则该订单为优质订单。
本发明的上述技术方案实现了非活跃司机的筛选，针对非活跃司机的优质订单推送，从而达到如下技术效果：
1.确保优质订单向非活跃司机的排他性播放，从而提高非活跃司机使用叫车软件的用户体验，提升司机使用叫车软件的留存率；
2.通过非活跃司机的订单播放，促使非活跃司机积极参与订单分配，也提升了出租车的个体运力和总体运力。
应用场景
设任一司机A，遍历叫车订单历史信息集合和出租车信息集合，查询是否有司机A的记录；
如果司机A在出租车信息集合中最近T1时间段内都有记录，并且司机A在叫车订单历史信息集合中最近T1时间段内没有记录，则司机A为非活跃司机；
将司机A的信息记录到所述非活跃司机集合。
作为一个示例，T1时间段为1～5天，例如：T1＝3天，则司机A为最近3天在线且不抢单的司机，说明司机A虽然经常在线，但抢单不积极，属于非活跃司机的一种情况。
所述非活跃司机判定模块另一种实施方式是：
设任一司机A，遍历出租车信息集合，查询是否有司机A的记录；
如果司机A在出租车信息集合中最近T2时间段内没有记录，并且司机A在出租车信息集合中最近T3时间段内有记录，则司机A为 非活跃司机；
将司机A的信息记录到所述非活跃司机集合。
则司机A最近3天不在线且最近30天内在线，说明司机A上线不积极，也属于非活跃司机的一种情况。
作为一个示例，T2时间段为1～5天，T3时间段内为15～45天，例如：T2＝3天，T3＝30天，则司机A最近3天不在线且最近30天内在线，说明司机A上线不积极，也属于非活跃司机的一种情况。
所述非活跃司机判定模块再一种实施方式是：
设任一司机A，遍历叫车订单历史信息集合，查询是否有司机A的记录；
设叫车订单历史信息集合在最近T4时间段内全体司机平均接单数量为DX，司机A在叫车订单集合中最近T4时间段内有D条记录，如果D小于等于DX的十分之一至二分之一(1/10～1/2)，则司机A为非活跃司机；
将司机A的信息记录到所述非活跃司机集合。
作为一个示例，T4时间段为1～5天，例如：T4＝3天，设叫车订单历史信息集合在最近3天内的平均订单数量为100，则如果司机A最近3天内接单数量为10，司机A接单数量等于各司机平均接单数量的十分之一，说明司机A接单意愿远低于各司机平均接单数量，司机A也属于非活跃司机的一种情况。
路线障碍物判断
实时路况常常存在复杂性，如，长期或短期存在的障碍物，如江河湖海、塌陷路面、正在施工的道路，都对交通服务的提供造成了困难。但目前，通常的交通信息服务难以考虑到以上信息，这样就带来了一些问题，对中国南方的某些城市如武汉、上海、杭州等，存在跨江接单的情况。以武汉为例，武汉市中心被长江切为两半，而现有的打车软件在计算司机与订单的距离时只能计算出直线距离，在这种情况下，某一订单A起点位置和某一司机甲当前位置分布在长江的两岸，直线距离只有0.5km，而司机甲从其当前位置实际接到订单A的 乘客则需要先行驶较远距离，过桥，再返回一段路程，实际行程可能达到5km，因此司机甲实际上不属于订单A起点位置周围的司机，不应当将订单A推送给司机甲。
本实施例能在叫车订单分配或推送过程中，克服只通过直线距离获取一订单周围的司机的方法的缺陷，避免向需要跨江、跨河以及跨越立交桥等障碍物的司机推送订单。
跨越障碍物判定
在本实施例中，处理模块具体的功能为判断各在线出租车针对某一叫车订单是否需要跨越障碍物。
优选的，所述处理模块包括如下步骤：
步骤1，采集障碍物上至少两个点，形成一个以上障碍物线段，取其中一个障碍物线段为P1P2；
步骤2，取叫车订单集合中任一叫车订单Q，设Q的起点位置为P3，设有一个以上的出租车符合出租车当前位置在叫车订单Q起点位置周围的条件，形成出租车信息集合；
步骤3，取出租车信息集合中任一个出租车C，设C的当前位置为P4，形成司机订单线段P3P4；
步骤4，判断P1P2和P3P4是否相交，如果是，则出租车C针对叫车订单Q需要跨越障碍物，转步骤6；如果否，转下一步骤；
步骤5，各障碍物线段是否已经遍历完，如果否，返回步骤4，取下一个障碍物线段判断是否与司机订单线段P3P4相交；如果是，则出租车C针对叫车订单Q不需要跨越障碍物，转步骤6；
步骤6，出租车信息集合是否已经遍历完，如果否，返回步骤3，取出租车信息集合中下一个出租车；如果是，转下一步骤；
步骤7，叫车订单集合是否已经遍历完，如果否，返回步骤2，取叫车订单集合中下一叫车订单；如果是，结束。
进一步，所述判断P1P2和P3P4是否相交的步骤是：
设定条件1为
Figure PCTCN2015086075-appb-000006
或者
Figure PCTCN2015086075-appb-000007
Figure PCTCN2015086075-appb-000008
设定条件2为
Figure PCTCN2015086075-appb-000009
或者
Figure PCTCN2015086075-appb-000010
Figure PCTCN2015086075-appb-000011
如果条件1和条件2同时满足，则P1P2和P3P4相交。
进一步，如果条件1满足，条件2不满足，但
Figure PCTCN2015086075-appb-000012
时，如果叫车订单Q的起点位置P3在P1P2线段上，P1P2和P3P4是相交的；如果叫车订单Q的起点位置P3在P1P2线段的延长线上，P1P2和P3P4是不相交的。
作为一个示例，所述障碍物为河流、湖泊、湿地或立交桥。
本发明的上述技术方案基于线段是否相交的判断，能检测出订单周围的司机是否有需要跨越障碍物的情形，达到如下技术效果：
1、在算法的性能和复杂度上有非常好优势，相比与传统的基于路面距离的计算，无需考虑实时路况，大大降低了服务器的计算性能；
2、避免向需要跨越障碍物的司机推送订单，也提高了订单推送的效率和准确度。
应用场景
如图18所示为处理模块判断路线障碍物流程图。
所述处理模块判断过程包括如下步骤：
步骤S1300，采集障碍物上至少两个点，形成一个以上障碍物线段，起点为P1，终点为P2，取其中一个障碍物线段为P1P2；
步骤S1310，取叫车订单集合中任一叫车订单Q，设Q的起点位置为P3，设有一个以上的出租车符合出租车当前位置在叫车订单Q起点位置周围的条件，形成出租车信息集合；
步骤S1320，取出租车信息集合中任一个出租车C，设C的当前位置为P4，形成司机订单线段P3P4；
步骤S1330，判断P1P2和P3P4是否相交，如果是，则进入步骤S1340，确定出租车C对于叫车订单Q需要跨越障碍物P1P2；如果否，则进入步骤S1350，确定出租车C对于叫车订单Q不需要跨越障碍物P1P2。可以采用基于向量的叉积(cross product)方式判断两线段是否相交，也可以采用其他方法判断两线段是否相交。
图19为判断两线段是否相交示意图。
采用向量的叉积(cross product)方式判断两线段对应向量
Figure PCTCN2015086075-appb-000013
和
Figure PCTCN2015086075-appb-000014
是否相交的原理是：如图16所示，设有线段P1P2，P3P4。
Figure PCTCN2015086075-appb-000015
和
Figure PCTCN2015086075-appb-000016
的叉积
Figure PCTCN2015086075-appb-000017
为正时，说明
Figure PCTCN2015086075-appb-000018
在
Figure PCTCN2015086075-appb-000019
的顺时针方向上
Figure PCTCN2015086075-appb-000020
为负时，说明
Figure PCTCN2015086075-appb-000021
在
Figure PCTCN2015086075-appb-000022
的逆时针方向上；叉积为0说明两向量共线(同向或反向)。因此，当同时满足以下两个条件时则
Figure PCTCN2015086075-appb-000023
和
Figure PCTCN2015086075-appb-000024
相交：
(1)
Figure PCTCN2015086075-appb-000025
和
Figure PCTCN2015086075-appb-000026
在
Figure PCTCN2015086075-appb-000027
的两侧(即
Figure PCTCN2015086075-appb-000028
或者
Figure PCTCN2015086075-appb-000029
)；
(2)
Figure PCTCN2015086075-appb-000030
和
Figure PCTCN2015086075-appb-000031
在
Figure PCTCN2015086075-appb-000032
的两侧(即
Figure PCTCN2015086075-appb-000033
或者
Figure PCTCN2015086075-appb-000034
)。
图20与图21是两种特殊情况下判断两线段是否相交的示意图。
图20和图21是特殊情况，满足条件(1)，不满足条件(2)，因为
Figure PCTCN2015086075-appb-000035
和
Figure PCTCN2015086075-appb-000036
重合，即
Figure PCTCN2015086075-appb-000037
和
Figure PCTCN2015086075-appb-000038
的叉积为0。当叉积为0时要分情况讨论，当P3在线段P1P2上时两线段相交；当P3在线段P1P2的延长线上时两线段不相交。
根据上述基于向量的叉积方式判断两线段是否相交的方法，所述判断P1P2和P3P4是否相交的步骤是：
设定条件(1)为
Figure PCTCN2015086075-appb-000039
或者
Figure PCTCN2015086075-appb-000040
Figure PCTCN2015086075-appb-000041
设定条件(2)为
Figure PCTCN2015086075-appb-000042
或者
Figure PCTCN2015086075-appb-000043
Figure PCTCN2015086075-appb-000044
如果条件(1)和条件(2)同时满足，则P1P2和P3P4相交。
进一步的，如果条件(1)满足，条件(2)不满足，但
Figure PCTCN2015086075-appb-000045
时，如果叫车订单Q的起点位置P3在P1P2线段上，此时P1P2和P3P4是相交的；如果叫车订单Q的起点位置P3在P1P2线段的延长线上，此时P1P2和P3P4是不相交的。
应用场景
以武汉为例，长江在武汉市内近似为一条直线，我们采用人工打点的方式，取长江在武汉的两端点(114.377997，30.666914)和 (114.157229，30.380211)。
某司机在湖北省武汉市江岸区中山大道1125号，经纬度(114.31108，30.604891)，某乘客在和平达到与秦园路交叉口发出订单，经纬度(114.338137，30.59485)，单看司机与乘客的直线距离在2km以内，而因为司机乘客联系横跨了长江，导致司机需要绕桥接乘客，实际距离近8km。
将上面长江两点(114.377997，30.666914)和(114.157229，30.380211)作为P1P2线段，司机(114.31108，30.604891)与订单(114.338137，30.59485)连线作为OD线段，根据上述基于向量的叉积方式判断两线段是否相交的方法可准确识别出P1P2线段与OD线段相交，因而判断该订单对该司机为跨江播送，进而将该订单对该司机隔离，即不向该司机播送该订单。
应用场景
上海的黄浦江相对复杂，在上海的走势不是一条直线，那么我们通过采集若干个关键点，将黄浦江在上海内的河段描述为一个折线，依次为P1P2、P2P3、P3P4、P5P6、P7P8，判断是否跨江播送只需判断司乘连线OD是否与上面的任一条线段相交即可。
以上所述实施例与应用场景示例仅是为充分说明本发明而所举的较佳的实施例，本发明的保护范围不限于此。本技术领域的技术人员在本发明基础上所作的等同替代或变换，均在本发明的保护范围之内。本发明的保护范围以权利要求书为准。
订单顺路判断
司机的行驶方向与订单是否顺路是影响司机和乘客用户体验的重要因素。在实际环境中，司机可能听到周边一定范围内的订单，但是经常会出现司机行驶方向和订单方位不一致的问题。诸如，司机当前正在向正东方向行驶，但是给司机播送的订单相对司机而言在正西方向。如果司机接了该订单，需要调头回去接乘客，在实际城市道路环境下，可能需要很长的时间(考虑允许调头路口、红绿灯、高架等)，严重影响司机和乘客的体验。
本实施例能实现在订单分配的排序阶段，将司机的行驶方向和订单方向之间的角度作为影响排序的因素，优先给司机发送司机行驶方向前方(顺路)的订单。
本实施例公开了一种处理订单的方法，包括通过处理模块：接收与订单相关的数据；获得服务提供者的运动方向；确定服务提供者的运动方向与从服务提供者位置到订单位置的方向之间的角度小于预定角度；以及将与订单相关的数据发送给服务提供者。
可选地，在该方法中，其中预定角度为90度。
可选地，在该方法中，还包括：在发送与订单相关的数据前，确定从承运方位置到订单位置的拥堵水平低于预定拥堵水平。
可选地，在该方法中，还包括：在发送与订单相关的数据前，确定从承运方位置到订单位置的实际行驶距离小于预定行驶距离。
可选地，在该方法中，还包括：在发送与订单相关的数据前，确定从承运方位置到预定位置的方向与从承运方位置到订单位置的方向之间的角度小于第二预定角度。
可选地，在该方法中，其中第二预定角度为90度。
可选地，在该方法中，还包括：在发送与订单相关的数据前，提升订单在订单列表中的排序。
根据本实施方式的原理，还公开了一种在承运方中处理订单的方法，包括：发送承运方的运动方向；以及在承运方的运动方向与从承运方位置到订单位置的方向之间的角度小于预定角度时，接收与订单相关的数据。
可选地，在该方法中，其中预定角度为90度。
根据本实施方式的原理，还公开了一种在承运方中呈现订单的方法，包括：接收与订单相关的数据；确定承运方的运动方向与从承运方位置到订单位置的方向之间的角度小于预定角度；以及呈现与订单相关的数据。
可选地，在该方法中，其中预定角度为90度。
可选地，在该方法中，还包括：在呈现与订单相关的数据前，确 定从承运方位置到订单位置的拥堵水平低于预定拥堵水平。
可选地，在该方法中，还包括：在呈现与订单相关的数据前，确定从承运方位置到订单位置的实际行驶距离小于预定行驶距离。
可选地，在该方法中，还包括：在呈现与订单相关的数据前，确定从承运方位置到预定位置的方向与从承运方位置到订单位置的方向之间的角度小于第二预定角度。
可选地，在该方法中，其中第二预定角度为90度。
可选地，在该方法中，还包括：在呈现与订单相关的数据前，提升订单在订单列表中的排序。
本实施例具有可以有效减少司机空驶里程和乘客等待时间，从而提升用户体验的优点。
应用场景
以下仅以出租车叫车服务为例对本发明的处理订单的方法和装置进行详细描述。
图22为判定顺路订单流程图。
根据图22所示，在步骤S1400中，在服务器接收与订单相关的数据。获取订单起点与终点的位置，计算起点与终点连线方向。
根据本实施例的各种示例性实施方式，本发明可以按照客户端-服务器架构被实施或者被实施在单个装置上。其中当按照客户端-服务器架构被实施时，可以由服务器来获得订单。获得订单的方式可以包括：通过信息接收模块直接从发出订单的服务请求者接收订单或者接收由其他中间机构(例如，某个网站等)转发的订单。在获得订单时获得与订单相关的数据，这些数据包括但不限于待乘坐出租车的人员等待出租车时所在的位置(简称为订单位置)、将要前往的目的地、愿意支付的额外小费、愿意等待的时间、乘坐人数、是否携带大件行李等。应当理解，上述待乘坐出租车的人员可以是使用打车软件来呼叫出租车的软件用户，也可以是该软件用户代为呼叫出租车的其他人员。订单位置可以由经纬度坐标表示，也可以在适当的情况下以其他可以用来表示确定的位置的信息表示，这些信息包括但不限于公交车 站、地铁站、某个路口以及某个特定建筑物等。当订单位置由除了全球定位系统坐标以外的信息表示时，可以由订单的接收方(例如，服务器)或者第三方(例如，诸如某个专业网站的其他地址解译机构)将其转换为全球定位系统坐标以便于进行后续操作。
继续参照图22，在步骤S1410中，获得承运方(司机)的运动方向。
根据本发明的各种示例性实施方式，承运方的运动方向由承运方例如通过全球定位系统提供。其中当按照客户端-服务器架构实施本发明时，可以由服务器通过直接从承运方获得承运方的运动方向或者通过其他中间机构(例如，全球定位系统信息提供机构)获得承运方的运动方向。应当理解，如上所述所获得的承运方的运动方向通常不是单独的方向信息，而是可以包括诸如承运方位置(例如，用全球定位系统坐标表示的位置)、承运方的运动速度等其他信息。因此，当需要在其他步骤中使用诸如承运方位置或运动速度等其他信息时，可以采取与以上所述的直接从承运方获得或者通过其他中间机构获得的方式来获得所需的信息。
继续参照图22，在步骤S1420中，在服务器确定承运方的运动方向与从承运方位置驶向订单位置的方向(简称为订单方向)之间的角度小于预定角度。
根据本发明的各种示例性实施方式，在步骤S1420中，首先确定订单方向。由于如上所述在步骤S1400和S1410中分别可以获得订单位置和承运方位置，因此可以通过获得用全球定位系统坐标表示的订单位置和承运方位置的经度和纬度坐标来计算出订单方向。
在计算出订单方向后，可以进一步计算出承运方的运动方向与订单方向之间的角度。图23图示了这一过程的详细步骤。
图23是司机运动方向与订单方向示意图。
参照图23，以被标记为N(北)的方向600作为参考。
首先，计算出司机(承运方)的运动方向610与N之间的角度α。而后，计算出订单方向620与N之间的角度β。最后，利用α和β计 算出承运方的运动方向610与订单方向620之间的角度θ。其中，θ可以被如下计算出：
Figure PCTCN2015086075-appb-000046
其中，由于0<α,β<2π并且0≤θ≤π，因此在计算θ时需要进行上述转化。计算出的θ的大小将在0和π之间，即在0度到180度之间。
在计算出运动方向与订单方向之间的角度θ之后，服务器确定角度θ是否小于预订角度。其中，当θ为0度时，说明订单位置恰好在司机(承运方)的运动方向610上，此时在不考虑实际道路的情况下承运方在去往订单位置时最为顺路，因此这一订单方向对于承运方而言是最佳订单方向。当θ在0度到90度之间时，说明订单方向与承运方的运动方向610大致同向，此时在不考虑实际道路的情况下承运方在去往订单位置时大致顺路。当θ为90度时，说明订单方向与承运方的运动方向610垂直，此时在不考虑实际道路的情况下承运方在去往订单位置时既不顺路也不绕路。当θ大于90度时，说明订单方向与承运方的运动方向610大致反向，此时在不考虑实际道路的情况下承运方在去往订单位置时大致绕路。因此，只有当θ小于90度时，承运方在去往订单位置时才能够顺路或者大致顺路，所以可以将预定角度设置为90度。应当理解，根据本发明的其他实施方式，在承运方对顺路有更高的要求时，可以将预定角度设置为更小的角度，诸如45度或者更小。
根据本发明的实施方式，当服务器确定承运方的运动方向610与订单方向之间的角度不小于预定角度时，服务器可以结束针对该承运方的对这一订单的处理。
根据本发明的各种示例性实施方式，预定行驶距离可以是具体距离(诸如4公里)，也可以是以从承运方位置到订单位置的直线距离为参照的距离(诸如该直线距离的1.5倍)。
继续参照图22，在步骤S1430中，处理模块计算订单起点与终点连线方向与司机运动方向的夹角小于夹角阈值。若是，则判断该订 单为该司机的顺路订单S1450；若否，则判断该订单为该司机的非顺路订单S1440。
图24是处理模块顺路订单判定、显示流程示意图。
如图24所示，在步骤S1500，处理模块获取订单信息，其中包括起点与终点的位置，计算订单的位移方向；
步骤S1510，处理模块获取司机的运动方向；
步骤S1520，处理模块计算订单位移方向与司机运动方向的夹角；
步骤S1530，处理模块判断所述夹角是否小于夹角阈值，若是，则判断该订单为该司机的顺路订单S1550，为该司机显示该订单信息S1560；若否，判断该订单为该司机的非顺路订单S1540。
抢单概率判断
使用打车软件的司机和乘客数量日益增多，如何实现对同时在线的大规模订单和司机进行快速的最佳匹配，对算法和架构是一个极具挑战性的问题。最佳匹配是指在考虑了订单特征、司机特征、周围司机量、时间、路况等诸多因素后，对每个司机呈现(例如，通过语音播送或者画面显示)了当前合适的订单，并且每个订单都得到了充分的呈现次数。
但是，作为示例，考虑在司机在线数目和司机在线时间一定的假设情形，司机能够听到的订单播送数是一定的。如果对司机播送订单不具备精准性，即不做区分地播送订单，不仅浪费宝贵的订单播送信道，使得司机愿意选择(即愿意抢单)的订单没有得到充分的播送，而且也对司机造成干扰。
本实施例可以在长期运营的过程中在存储模块存储每日的订单呈现日志和司机抢单日志。运用机器学习和数据挖掘的方法，利用这些海量的日志对模型进行训练，可以应用模型从而精准地预估出关于司机选择每次呈现的订单的抢单概率。将预估出来的司机抢单概率用于订单分配。
根据本实施例，提供了一种处理订单的方法，包括：获取历史订单的至少一个特征以及与历史订单相关联的服务提供者的响应；根据 与历史订单相关联的服务提供者的响应，向至少一个特征分配权重；获取当前订单的特征；以及根据与当前订单的特征相对应的权重，选择当前订单中的将向服务提供者呈现的当前订单。
可选地，在该方法中，获取历史订单的至少一个特征以及与历史订单相关联的服务提供者的响应包括：获取服务提供者是否选择历史订单的响应。
可选地，在该方法中，根据与历史订单相关联的服务提供者的响应，向至少一个特征分配权重包括：利用机器学习模型，根据历史订单的至少一个特征以及服务提供者是否选择历史订单的响应，向至少一个特征分配权重。
可选地，在该方法中，根据与当前订单的特征相对应的权重，选择当前订单中的将向服务提供者呈现的当前订单包括：利用机器学习模型，根据当前订单的特征以及与当前订单的特征相对应的权重，确定服务提供者选择当前订单的概率。
可选地，在该方法中，根据与当前订单的特征相对应的权重，选择当前订单中的将向服务提供者呈现的当前订单还包括：选择当前订单中的被服务提供者选择的概率最高的当前订单，作为将向服务提供者呈现的当前订单。
可选地，在该方法中，机器学习模型包括逻辑回归模型或者支持向量机模型。
可选地，在该方法中，还包括：利用将向服务提供者呈现的当前订单的特征以及与当前订单相关联的服务提供者的响应，更新与当前订单的特征相对应的权重。
可选地，在该方法中，特征包括以下各项中的至少一项：发送订单的位置与服务提供者的距离、订单的目的地、订单的目的地种类、订单的目的地周围的路况以及订单的呈现次数。
可选地，在该方法中，历史订单、当前订单以及服务提供者与同一地理区域相关联。
图25是处理模块处理订单的方法的流程图
根据本发明的实施方式，可以在机器学习过程的训练阶段期间或者之前执行步骤S1600或S1610。可以在在线服务器中或者在大数据平台服务器中执行这些步骤。具体而言，可以于在线服务器中获取日志并将其存储为历史订单-服务提供者标识配对的形式作为总体样本。可选地，可以在另一服务器中，例如大数据平台服务器，从在线服务器获取日志并存储为历史订单-服务提供者标识配对的形式作为总体样本。
步骤S1600，获取历史订单的至少一个特征以及与历史订单相关联的服务提供者响应；历史订单的至少一个特征可以包括：订单的费用，愿意支付的额外小费，发送订单的位置与服务提供者的距离，乘坐人数，或者待乘坐出租车的人员等待出租车时所在的位置与服务提供者的距离、订单中将要前往的目的地，订单目的地的类型(例如，机场、火车站、医院或者学校)、订单的目的地周围的路况或者订单的呈现次数。历史订单的特征还可以包括：愿意等待的时间、是否携带大件行李、订单的事由等；
步骤S1610，获取服务提供者是否选择历史订单的响应，该响应包括，抢单行为及其结果，如是否抢单，抢单成败，发出抢单命令距离接收到订单的时间(表征抢单速度和接单意愿)等；
步骤S1610本身不是必需的，也可以从整个流程中删掉。
在步骤S1600和S1610之后执行步骤S1620，根据与历史订单相关联的服务提供者的响应，向至少一个特征分配权重；
特别地，在步骤S1620后，可以执行一个步骤S1630，在此，利用机器学习模型，根据历史订单的至少一个特征以及服务提供者是否选择历史订单的响应，向至少一个特征分配权重。这一步骤也不是必需的，可以直接向特征分配权重。
在某些实施方式中，机器学习模型可以是广泛运用于二分类问题的逻辑回归(Logistic Regression)模型。在某些实施方式中，机器学习模型可以是支持向量机模型。在其它实施方式中，还可以根据测试结果来使用其它的机器学习模型。
考虑使用逻辑回归模型的情况。假设历史订单的特征、或者预测变量X＝x，则服务提供者是否选择历史订单的响应的布尔表示，即目标变量Y＝1的概率如下公式表示：
Figure PCTCN2015086075-appb-000047
其中β为向特征分配的权重，或者称为模型参数。
在训练样本中，如果服务提供者选择历史订单，则Y＝1；如果服务提供者不选择历史订单，则Y＝0。基于确定的预测变量X以及目标变量Y，将包括训练数据中的多组历史订单-服务提供者配对的信息的预测变量X和目标变量Y代入公式，能够确定向至少一个特征分配的权重β，β可以是与预测变量X的向量相对应的向量。换句话说，每个特征可以具有对应于该特征的一个权重。作为示例，位置特征可以具有权重0.5，路况特征可以具有权重0.3，并且订单呈现次数特征可以具有权重0.1。
根据某些实施方式，在向特征分配的权重之后，可以将向特征分配的权重存储在数据文件中。当在大数据平台服务器中进行训练的情况下，可以将数据文件发送至在线服务器以便在线服务器在应用阶段加载该数据文件以获取权重。
在步骤S1630后，进入步骤S1640，获取当前订单的特征；
当前订单是指有待向服务提供者呈现或者正在呈现的订单。例如，当前订单可以是尚未向服务提供者呈现的订单，或者是正在向一些服务提供者呈现而尚未向其他服务提供者呈现的订单。可以由在线服务器中获取当前订单。获取订单的方式可以包括直接从发出订单的待乘坐出租车的人员接收订单或者接收由其他中间机构(例如，某个网站等)转发的订单。
在获取当前订单之后，可以在众多服务提供者中选择与当前订单相关联的一个或多个服务提供者作为将向其呈现当前订单的候选服务提供者。作为示例，可以选择处于发送当前订单的位置某范围内的服务提供者作为候选服务提供者。还可以根据利用其它因素，例如服务提供者的行驶方向等来选择候选服务提供者。另外，还可以对所选 择的候选服务提供者进行进一步的过滤。
注意到，在获取多个当前订单之后，可以有多个当前订单与单个服务提供者相关联，例如在该服务提供者处于多个当前订单的发送位置的特定范围内的情况下。因而，可以选择多个当前订单中的一个优选订单向该服务提供者呈现。
当前订单的特征可以是与在训练阶段向其分配权重的特征相对应的特征。因而，当前订单的特征可以也包括：发送订单的位置与服务提供者的距离或者待乘坐出租车的人员等待出租车时所在的位置与服务提供者的距离、订单中将要前往的目的地，订单的目的地种类(例如，机场、医院或者学校)、订单的目的地周围的路况或者订单的呈现次数。当前订单和服务提供者的特征还可以包括：愿意支付的额外小费、愿意等待的时间、乘坐人数、是否携带大件行李等。此外，如关于历史订单进行描述的，当前订单的特征可以是从当前订单所确定的内容直接确定的，或者可以是利用服务器对所确定的内容进行处理而进一步间接确定的。针对当前订单的特征，可以利用与当前订单的特征相对应的权重，即在训练阶段向与当前订单的特征相对应的特征分配的权重。
在步骤S1650，可以根据与当前订单的特征相对应的权重，选择当前订单中的将向服务提供者呈现的当前订单；
在步骤S1660，根据当前订单的特征以及与当前订单的特征相对应的权重，确定服务提供者选择当前订单的概率，这一步骤也不是必需的；
根据本实施方式，在获取当前订单的特征之后，可以将当前订单和与当前订单相关联的候选服务提供者所组成的当前订单-服务提供者配对中的特征的信息组成向量以作为预测变量，并且将预测变量以及以向量表示的、与预测变量中的特征相对应的权重一起代入机器学习模型进行应用。利用机器学习模型，根据预测变量和对应的权重，能够确定服务提供者选择当前订单的概率。在某些实施方式中，在应用阶段使用的机器学习模型可以是与在训练阶段相同的机器学习模 型，例如逻辑回归模型。在某些实施方式中，机器学习模型可以是支持向量机模型。在其它实施方式中，还可以根据测试结果来使用其它的机器学习模型。仍考虑使用逻辑回归模型的情况。将在当前订单的特征、或者预测变量X＝x，以及对应的权重β代入在训练阶段示出的公式。则能够得出服务提供者是否选择当前订单的概率，即目标变量Y，其中Y可以是从0至1之间的实数。
接着，可选地，如果服务提供者选择当前订单的概率高于预定阈值，则可以将当前订单加入订单列表以用于排序。根据本实施方式，在线服务器可以将由服务器在不同时间接收的针对一个服务提供者的多个当前订单组成订单列表以用于排序。将该服务提供者选择当前订单的概率与预定阈值进行比较，如果概率高于预定阈值，则将当前订单加入订单列表以用于排序；如果概率低于预定阈值则不加入订单列表。通过这一步骤，可滤除服务提供者明显不希望选择的当前订单。根据一些实施方式，可以将预定阈值存储在在线服务器中的相应程序的配置文件中。还可以根据需要，基于服务提供者的响应、订单在特定时间的整体匹配状况等因素来定期或者动态地调整该阈值。在一些实施方式中，可以不进行前文所述的选择与当前订单相关联的候选服务提供者，而直接通过概率来筛选与当前订单相关联的服务提供者。
在步骤S1660后，接着执行步骤S1670。
在步骤S1670中，可以选择当前订单中的被服务提供者选择的概率最高的当前订单，作为将向服务提供者呈现的当前订单。根据本发明的实施方式，根据一些实施方式，在线服务器可以根据服务提供者选择多个当前订单的概率，对这些当前订单进行排序，然后选择被服务提供者选择的概率最高的当前订单作为将向服务提供者呈现的当前订单。
根据一些实施方式，在选择将向服务提供者呈现的当前订单之后，可以将当前订单发送给服务提供者的客户端以向服务提供者呈现。例如，可以在服务提供者的安装有客户端的设备(例如，移动设备)上，通过语音播送或者在用户接口(例如，触敏显示器等)上以 画面显示的方式来呈现当前订单。服务提供者可以取决于其是否对该当前订单感兴趣，而通过用户接口选择该当前订单作为响应。可选地，根据一些实施方式，在线服务器可以将多个当前订单与服务提供者选择当前订单的概率一起发送给服务提供者，并且在服务提供者的安装有客户端的设备上实现多个当前订单的排序。根据以上所描述的内容可见，可以呈现具有更高的被服务提供者选择的概率的订单在，从而实现优先呈现服务提供者更愿意选择的订单。
根据一些实施方式，在服务提供者选择当前订单之后，或者服务提供者未选择当前订单从而订单被呈现预定时间之后，可以返回步骤S1640，以获取新的当前订单以用于选择；或者返回步骤S1650或S1660，基于部分更新的特征，重新选择将向服务提供者呈现的订单。注意到，同一当前订单在其有效期间，可能由于较高的被选择的概率而向同一服务提供者多次呈现。就此而言，可以记录同一当前订单向同一服务提供者重复呈现的次数作为该当前订单的特征。
在步骤S1680，利用将向服务提供者呈现的当前订单的特征以及与当前订单相关联的服务提供者的响应，更新与当前订单的特征相对应的权重。在步骤S1680后，可以回到步骤S1620，根据与历史订单相关联的服务提供者的响应，向至少一个特征分配权重。
基于订单长度的抢单概率判断
本实施例，具体地，针对订单长度预估抢单概率。该方法包括：获取当前订单的始发地与目的地之间的距离；获取该服务提供者对于历史订单的抢单概率，其中该历史订单的始发地与目的地之间的距离和该当前订单的始发地与目的地之间的距离相关；以及基于该抢单概率，向该服务提供者发送该当前订单。
这样，能够减少该服务提供者抢单概率较低的订单的发送，即能够减少针对该服务提供者来说无价值或低价值的订单的发送，从而保证针对该服务提供者来说高价值的订单的快速地、精准地发送。
如上文已经详细描述的，该服务提供者既可以包含传统意义驾驶车辆、船、飞行器的司机，也可以包含无人驾驶时用于载客/载物的交 通工具。
根据本实施例，首先可以分别获取当前订单的始发地与服务提供者的位置，然后计算该当前订单的始发地与该服务提供者的位置之间的距离。其中，该当前订单的始发地可以从上述订单信息中来获取；该服务提供者的位置可以经由该服务提供者的智能设备中的定位信息来确定。另外，对于该当前订单的始发地与该服务提供者的位置之间的距离，既可以是它们之间的直线距离，也可以是当它们被置于导航系统中时参考路线信息、路况信息和路政信息而计算得到的车辆实际行驶距离。接下来，获取该服务提供者对于历史订单的抢单概率，其中该历史订单的始发地与该服务提供者之间的距离和该当前订单的始发地与该服务提供者之间的距离相关。例如，当存在大量历史订单使得每个历史订单的始发地与该服务提供者之间的距离和该当前订单的始发地与该服务提供者之间的距离都相关时，既可以分别获取该服务提供者对于每个历史订单的抢单概率，也可以获取该服务提供者对于这些历史订单整体的抢单概率。
根据本实施例，该历史订单的始发地与该服务提供者之间的距离和该当前订单的始发地与该服务提供者之间的距离的相关性可以体现如下：(1)该历史订单的始发地与该服务提供者之间的距离等于该当前订单的始发地与该服务提供者之间的距离。
在这一实施例中，因为该历史订单的始发地与该服务提供者之间的距离等于该当前订单的始发地与该服务提供者之间的距离，所以该服务提供者对于该历史订单的抢单概率在很大程度上等于该服务提供者对于该当前订单的抢单概率。也就是说，如果该服务提供者对于该历史订单的抢单概率较低，则该服务提供者对于该当前订单的抢单概率在很大程度上也较低。这样，该当前订单对于该服务提供者来说将可能是无价值或低价值的，因此该当前订单的发送将可能影响针对该服务提供者来说高价值的订单的发送。因此，在这一实施例中，通过减少该服务提供者抢单概率较低的当前订单的发送，能够保证针对该服务提供者来说高价值的订单的快速地、精准地发送。(2)该历史 订单的始发地与该服务提供者之间的距离和该当前订单的始发地与该服务提供者之间的距离属于相同的距离区间，其中该距离区间按照各个历史订单的始发地与服务提供者之间的距离而预先分配。例如，0-100米是第一距离区间，通过P1来表示；100-200米是第二距离区间，通过P2来表示；200-300米是第三距离区间，通过P3来表示；以此类推。
在这一实施例中，因为该历史订单的始发地与该服务提供者之间的距离和该当前订单的始发地与该服务提供者之间的距离属于相同的距离区间，所以该服务提供者对于该历史订单的抢单概率在很大程度上近似于该服务提供者对于该当前订单的抢单概率。也就是说，如果该服务提供者对于该历史订单抢单概率较低，则该服务提供者对于该当前订单的抢单概率在很大程度上也较低。这样，该当前订单对于该服务提供者来说将可能是无价值或低价值的，因此该当前订单的发送将可能影响针对该服务提供者来说高价值的订单的发送。因此，在这一实施例中，通过减少该服务提供者抢单概率较低的当前订单的发送，能够保证针对该服务提供者来说高价值的订单的快速地、精准地发送。
图26是处理订单的方法的流程图。根据本实施例，首先，在步骤S1700，获取当前订单的始发地与目的地之间的距离。执行步骤S1710，确定当前订单的始发地与服务提供者之间的距离，将该距离归属于一个订单距离区间所属的距离区间；然后可以执行步骤S1720，获取一个司机对于处于该订单距离区间的历史订单的抢单概率，这一概率通常可以从推送给该司机的历史订单与该司机对于历史订单的响应记录中求解得到。
接下来在步骤S1730，判断该抢单概率是否大于某一概率阈值，若否，进入步骤S1740，不为该司机推送该订单信息；若是，进入步骤S1750，为该司机推送该订单信息。
图27是根据本实施例的用于生成抢单概率向量流程流程图，包括如下的步骤S1800至步骤S1840。
步骤S1800，对于每个服务提供者，处理服务提供者的订单信息，获取例如近一个月内接收到的历史订单的播单距离和抢单信息，其中播单距离是指订单的始发地与服务提供者的位置之间的距离；
步骤S1810，将多个历史订单按照距离区间归属于多个订单集合；
步骤S1820，计算针对不同播单距离的抢单概率。例如，对于0-100米的播单距离、100-200米的播单距离以及200-300米的播单距离，分别计算抢单概率。其中，作为一个示例，该对于0-100米的播单距离的抢单概率可以等于近一个月内接收到的播单距离在0-100米中的历史订单的抢单次数与近一个月内接收到的播单距离在0-100米中的历史订单的接收次数的百分比。
步骤S1830，生成抢单概率向量，作为一个示例，假设距离区间分别为(A，B，C，D，E，F)，A区间为0-10km，B为10-20km，C为20-30km，D为30-40km，E为40-50km，F为50-60km，抢单概率向量可以是(0.1，0.15，0.2，0.4，0.3，0.2)，这一向量表示司机对于距离为30-50km的订单最感兴趣；
在步骤S1840，存储该抢单概率向量。
以上所述为处理模块能实现的一些对于订单的判断模式与方法，在实际应用中并不局限于上述方法与判断模式，以上举例仅为说明之用，并不限制本发明的保护范围。
订餐服务实施例
现有的订餐分配系统，往往需要订餐者首先确定某个目标餐馆，并进入该餐馆的订餐界面进行操作。这需要订餐者对餐馆有十分清楚的了解，对餐饮需求比较明确。但实际情况中，常常存在订餐者对周围环境不熟悉、对餐馆的经营情况、消费水准、经营菜系不了解等状况。
本实施例提供了一种基于分析订餐订单信息和餐馆信息，选取目标餐馆，并向该目标餐馆发送订餐订单的系统。其中，订餐订单信息包括但不限于，订餐者姓名、订餐电话、送餐地址、送餐时间、菜系、可以接受的价位区间、订餐者的偏好口味等；来自餐馆的信息包括餐 馆营业时间、餐馆地址、餐馆联系电话、餐馆经营的餐饮类型、通常的消费区间等。
其中订单提交的菜系包括但不限于川菜、粤菜、鲁菜、粤菜、闽菜、苏菜、淮扬菜、浙菜、湘菜、徽菜、法国菜、意大利菜、土耳其菜、墨西哥菜、日式料理、韩式料理等。其中，订餐者在订餐订单中可以选择一个确定的菜，也可以是只给出自己想要吃的类型，系统依照订餐者的历史订单记录及喜好口味，选取目标餐馆可以是一种或多种特定的食物，也可以只是一个菜系概述，由系统依照订餐者过往的订单，偏好等为订餐者推荐特定食物。并结合其他订单信息，选择目标餐馆，发送订单。
图28是目标餐馆判定流程图。
在本实施例中，在步骤S1900，取订餐订单集合中一订餐订单Q，取餐馆信息集合中一餐馆信息C；
在步骤S1910，信息模块110获取订餐订单的送餐地址D1，送餐时间T1，餐饮类型y，接受价位区间为X1，获取餐馆信息C中餐馆的营业时间[T2，T3]，餐馆地址D2，餐馆提供的餐饮类型集合为Y，餐馆消费价位区间为X2，距离阈值D；
在步骤S1920，处理模块130接收信息模块发送的信息，并作进一步的判断处理。首先判断餐馆地址D2与送餐地址D1之间的距离是否小于或等于距离阈值D(|D2-D1|≤D)，并满足订餐者要求的订餐时间T1是处于餐馆的营业时间范围内(T2≤T1≤T3)，如果上述两个条件均满足，则进入步骤S1930继续判断，如果其中有一项不满足，则进入步骤S1960，判定该餐馆不是目标餐馆；
在步骤S1930，判断订餐者想要的餐饮类型y是否属于餐馆可提供的餐饮类型集合Y(y∈Y)，如果属于，则进入步骤S1940，如果不属于，则进入步骤S1960，判定该餐馆不是目标餐馆；
在步骤S1940，判断订餐订单的接受的价位区间X1与餐馆消费价位区间X2是否存在交集，如果存在交集，进入步骤S1950，则判定该餐馆为目标餐馆；如果没有交集
Figure PCTCN2015086075-appb-000048
则进入步骤S1960， 判定该餐馆不是目标餐馆。
需要注意的是步骤S1920，S1930，S1940的顺序可以颠倒变换，同时这几个步骤不是全都必需的，可以根据需要，删除一个或者两个判断步骤，而不影响在后续步骤中对订餐订单的目标餐馆的判断。例如，可以只判断S1920中的条件是否满足，而决定餐馆C是否是订餐订单Q的目标餐馆，而不再进行S1930和/或S1940中条件的判断。
在进入步骤S1950，判断该餐馆为目标餐馆后，系统可以进一步地向目标餐馆发送订餐订单的信息。
应用场景
订餐者首先通过自己的手机客户端，采用手写输入、语音输入、键盘输入的方式输入自己的订餐请求信息，如，订餐者：用户A，送餐时间：中午11:30，送餐地址：北京市海淀区中关村海龙电子商城5层501室，参考价位：15-25元，餐饮类别：快餐，偏好：微辣，从而在订餐软件的服务器上生成具有上述信息的订单。
Figure PCTCN2015086075-appb-000049
订餐软件的服务器上存有大量的餐馆信息，截取了其中两家餐馆信息，如下所示：
Figure PCTCN2015086075-appb-000050
服务器基于当前的订单信息和餐馆信息，会对订单和餐馆进行匹配，根据目标餐馆的判断规则，餐馆0012对于订单12001而言属于目标餐厅，向该餐厅推送该订餐订单。
送货服务实施例
本实施例提供了一种基于分析快递订单信息和快递员信息，选取目标快递员，并向该目标快递员发送快递订单的系统。其中，快递订单信息包括订单起点位置、发快递时间、终点位置、客户等待时间阈 值等；来自快递员的信息包括当前位置、快递员偏好的送件距离区间、快递员到订单起点位置预估时间、接订单距离阈值等。
图29是目标快递员判定流程图。
在步骤S2000，取信息接收模块所获得的快递订单信息集合中的任一快递订单Q，取信息接收模块所获得的快递员信息信息集合中任一快递员信息C；
在步骤S2010，从快递订单Q中获取起点位置D1，终点位置D2，发快递时间T1，客户等待时间阈值T，从快递员信息C中获取快递员当前位置D3，快递员偏好的送件距离区间[D4，D5]，快递员到订单起点位置预估时间T2，接订单距离阈值D；
在步骤S2020，处理模块接收来自信息模块的信息，进行判断处理，首先判断快递员当前位置D3与快递订单的起点位置D1的距离(记为|D3-D1|)是否小于接订单的距离阈值D(|D3-D1|≤D)，若小于或等于接订单的距离阈值，则继续步骤S2030，若大于接订单的距离阈值，则进入步骤S2060，判定该快递员为非目标快递员；
在步骤S2030，判断订单的接送距离(|D2-D1|)是否处于快递员偏好的送件距离区间(|D2-D1|∈[D4,D5])，如果落入该区间，则执行步骤S2040，如果不在该区间，则进入步骤S2060，判定该快递员不是目标快递员，流程结束；
其中，订单起点位置，可以是寄件人当前的位置，也可以是由外地派发到某一集散中心的地理位置；
在步骤S2040，判断快递员由当前位置到订单起点位置的预估时间T2与发快递时间T1之间偏差是否小于或等于客户等待时间阈值T(|T1-T2|≤T)，如果小于或等于客户等待时间阈值T，则进入步骤S2050，判定该快递员为目标快递员，流程结束；如果大于客户等待时间阈值T，则进入步骤S2060，判定该快递员为非目标快递员。
需要注意的是步骤S2020、S2030、S2040的顺序可以颠倒变换，同时这几个步骤不是全都必需的，可以根据需要，删除一个或者两个判断步骤，而不影响在后续步骤中对订单的目标快递员的判断。例如， 可以只判断S2020中的条件是否满足，而决定快递员C是否是快递订单Q的目标快递员，而不再进行S2030和/或S2040中条件的判断。
进入步骤S2050，判定该快递员为目标快递员后，系统可以进一步地向目标餐馆发送订餐订单的信息。
应用场景
寄件人首先通过自己的智能设备，进入应用界面，采用手写输入、语音输入或键盘输入的方式输入自己的快递请求信息，如，寄件时间：上午09:30，寄件地址：北京市海淀区中关村海龙电子商城5层501室，寄件人姓名：用户A，电话：123…45，收件地址：北京市海淀区学院路20号，从而在快递服务的服务器上生成具有上述信息的订单。
Figure PCTCN2015086075-appb-000051
快递服务的服务器上存有大量的快递员信息，截取了两位快递员信息，如下所示：
Figure PCTCN2015086075-appb-000052
服务器基于当前的订单信息和快递员信息，会对订单和快递员进行匹配，根据目标快递员的判断规则，快递员0012对于订单12001而言属于目标快递员，向该快递员推送该订餐订单。
以上通过具体实施例，对服务派发系统在交通服务、餐饮服务、送货服务中的具体应用进行了描述。需要注意的是，以上示例或列举，并不用于局限服务派发系统的应用领域或场景，本领域的技术人员，在理解了本申请的内容后，可以容易地将本发明的构思，应用到其他服务领域，例如，家政服务等。
因为可以在不背离权利要求所限定的系统和方法的情况下，利用上述特征的这些和其他变化和组合，因此，应当将示例性实施例的前面描述当作为说明而不是对权利要求所限定的主题的限制。还应当理 解的是，示例的提供(以及表述为“诸如”、“例如”、“包括”等的分句)不应当被解释为将所要求的主题限制在特定示例；相反，示例意在仅说明许多可能方面中的部分。除非明确说明为相反的，否则，可以在本文的任何其他实施例、替选或示例中使用给定实施例、替选或示例的每个特征。
如本说明书和贯穿下面的权利要求书中所使用的那样，“一”、“一个”、“一种”、“该”以及“所述”也旨在包括复数形式，除非上下文明确指示其他情形。此处使用了术语“包括”与“包含”旨在包括已标识的方法或设备中的步骤或元素，但是这样的步骤或元素不构成排它性的列表，方法或设备可以包含额外的步骤或元素。此外，如本文说明书和贯穿下面的权利要求书中所用的那样，“在…中”包括“在…中”和“在…上”，除非上下文明确指示其他情形。
应当指出，可以将示例实施例描述为体现为流程图、流程图表、数据流程图、结构图或者框图的过程。尽管流程图可以将操作描述为顺序的过程，可以并行地、并发地或者同时地执行许多操作。此外，操作的顺序可以被重排。当过程的操作完成时，过程可以被终止，但是过程也可以具有不包括在图中的额外的步骤。过程可以对应于方法、功能、程序、子例程、子程序等。当过程对应于功能时，过程的终止可以对应于功能对调用函数或者主函数的返回。
尽管在某些实施例对应的附图中，在功能上将处理器和存储器图示为在同一相应块内，然而，本领域的技术人员应该理解，上述的各装置、各单元或各步骤可以借助于通用的计算装置来实现，它们可以集中在单个的计算装置上，如作为一个服务派发设备；也可以不处于同一个装置内。例如，信息接收模块、存储模块和处理模块实际上可以包括不位于同一物理外壳内的多个具体的设备，如无线收发装置、存储器、处理器、微处理器等。根据本发明的系统中的各个模块，可以分布在多个计算装置所组成的网络上，该网络可以是通过有线连接的，也可以是在一个区域内通过无线连接的，还可以是在不同地域内通过分布式网络连接的。可选地，它们也可以用计算装置可执行的程 序代码来实现，从而可以将它们存储在存储装置中由计算装置来执行，或者将它们分别制作成各个集成电路模块，或者将它们中的多个模块或步骤制作成单个集成电路模块来实现。这样，本发明不限制于任何特定的硬件和软件结合。
以上所述仅为本发明的一部分的可选实施方式，仅为说明本发明的技术构思，其目的在于让技术人员能够了解本发明的内容并据以实施，并不用于限制本发明，对于相关领域的技术人员来说，本发明可以有各种更改和变化。凡是在本发明的精神和原则之内，所做出的任何修改、等效替换、改进等，均应包含在本发明的保护范围之内。

Claims (21)

  1. 一种服务派发系统，包含：
    一个信息接收模块，被配置为接收来自一个服务提供者的服务提供信息与来自一个服务请求者的服务请求信息；
    一个存储模块，被配置为存储所述信息接收模块接收到的服务提供信息与服务请求信息；
    一个处理模块，被配置为，将所述存储模块所存储的服务提供信息与服务请求信息进行计算，以得到一个特征结果，i)如果所述特征结果满足至少一个判据，确定向所述服务提供者派发所述服务请求信息，ii)如果所述特征结果不满足至少一个判据，确定不向所述服务提供者派发所述服务请求信息；
    其中，所述服务请求信息包含两个地理位置。
  2. 根据权利要求1所述的服务派发系统，进一步包含一个输出模块，被配置为当确定向所述服务提供者派发所述服务请求信息时，将所述服务请求信息提供给所述服务提供者。
  3. 根据权利要求1所述的服务派发系统，其中所述处理模块包含一个判据存储单元，被配制为存储所述至少一个判据。
  4. 根据权利要求1所述的服务派发系统，其中所述服务提供信息包含服务提供者的当前信息。
  5. 根据权利要求4所述的服务派发系统，其中所述服务提供者的当前信息包含所述服务提供者的定位与运动信息。
  6. 根据权利要求5所述的服务派发系统，其中所述服务提供者的定位与运动信息包含所述服务提供者的位置信息。
  7. 根据权利要求5所述的服务派发系统，其中所述服务提供者的当前信息包含所述服务提供者的速度信息。
  8. 根据权利要求7所述的服务派发系统，其中所述服务提供者的速度信息包含所述服务提供者的速度方向。
  9. 根据权利要求1所述的服务派发系统，其中所述信息接收模块进 一步被配置为接收来自一个信息源的信息。
  10. 根据权利要求1所述的服务派发系统，其中，所述至少一个判据选自由下列组成的群组：
    指示服务提供者对服务请求信息响应的参数；
    指示服务提供者的活跃程度的参数；
    服务提供者常驻的地理位置与所述两个地理位置的距离；
    所述两个地理位置所构成的矢量与服务提供者的速度方向的夹角。
  11. 一种服务派发方法，包含：
    在一个信息接收模块，接收来自一个服务提供者的服务提供信息与来自一个服务请求者的服务请求信息；
    通过一个存储模块，存储所述信息接收模块接收到的服务提供信息与服务请求信息；
    通过一个处理模块，将所存储的服务提供信息与服务请求信息进行计算，以得到一个特征结果，i)如果该特征结果满足至少一个判据，确定向所述服务提供者派发所述服务请求信息，ii)如果该特征结果不满足至少一个判据，确定不向所述服务提供者派发所述服务请求信息；
    其中，所述服务请求信息包含至少两个地理位置。
  12. 根据权利要求11所述的服务派发方法，进一步包含：
    当确定向所述服务提供者派发所述服务请求信息时，通过一个输出模块，将所述服务请求信息提供给所述服务提供者。
  13. 根据权利要求11所述的服务派发方法，其中所述服务提供信息包含服务提供者的当前信息。
  14. 根据权利要求13所述的服务派发方法，其中所述服务提供者的当前信息包含所述服务提供者的定位与运动信息。
  15. 根据权利要求14所述的服务派发系统，其中所述服务提供者的定位与运动信息包含所述服务提供者的位置信息。
  16. 根据权利要求14所述的服务派发系统，其中所述服务提供者的当前信息包含所述服务提供者的速度信息。
  17. 根据权利要求16所述的服务派发方法，其中所述服务提供者 的速度信息包含所述服务提供者的速度方向。
  18. 根据权利要求11所述的服务派发方法，进一步包含：
    在所述信息接收模块，接收来自一个信息源的信息。
  19. 根据权利要求11所述的服务派发方法，其中，所述至少一个判据选自由下列组成的群组：
    指示服务提供者对服务请求信息响应的参数；
    指示服务提供者的活跃程度的参数；
    服务提供者常驻的地理位置与所述两个地理位置的距离；
    所述两个地理位置所构成的矢量与服务提供者的当前速度方向的夹角。
  20. 根据权利要求11所述的服务派发方法，进一步包含：
    通过所述处理模块，为对所述服务请求信息分配积分。
  21. 根据权利要求20所述的服务派发方法，进一步包含：
    当所述服务提供者执行所述服务请求信息后，为所述服务提供者发放所述积分。
PCT/CN2015/086075 2014-08-04 2015-08-04 服务派发系统及方法 WO2016019857A1 (zh)

Priority Applications (7)

Application Number Priority Date Filing Date Title
EP15829451.2A EP3179420A4 (en) 2014-08-04 2015-08-04 Service distribution system and method
KR1020177003867A KR20180006871A (ko) 2014-08-04 2015-08-04 주문들을 분배하기 위한 방법들 및 시스템들
MYPI2017000173A MY188692A (en) 2014-08-04 2015-08-04 Methods and systems for distributing orders
SG11201700895YA SG11201700895YA (en) 2014-08-04 2015-08-04 Methods and systems for distributing orders
KR1020187037289A KR20190000400A (ko) 2014-08-04 2015-08-04 서비스 분배 시스템 및 방법
US15/501,824 US20170228683A1 (en) 2014-08-04 2015-08-04 Methods and systems for distributing orders
PH12017500192A PH12017500192A1 (en) 2014-08-04 2017-02-01 Methods and systems for distributing orders

Applications Claiming Priority (20)

Application Number Priority Date Filing Date Title
CN201410379713.3 2014-08-04
CN201410379713.3A CN104167093B (zh) 2014-08-04 2014-08-04 基于司机住址信息的顺风车运营系统
CN201410397679.2 2014-08-13
CN201410397679.2A CN104156443B (zh) 2014-08-13 2014-08-13 地理兴趣点模糊改写系统
CN201410409108.6A CN104183118B (zh) 2014-08-19 2014-08-19 基于拍卖模式获得乘客最优接驾司机的派单系统
CN201410409108.6 2014-08-19
CN201410413040.9 2014-08-20
CN201410413040.9A CN104157133B (zh) 2014-08-20 2014-08-20 基于司机在线活跃情况的运力拉升系统
CN201410418423.5A CN104156868A (zh) 2014-08-22 2014-08-22 基于订单价值判断促进订单成交的出租车积分系统
CN201410418423.5 2014-08-22
CN201410421805.3A CN104183123B (zh) 2014-08-25 2014-08-25 基于判断线段是否相交的叫车订单播送系统
CN201410421805.3 2014-08-25
CN201410437102.X 2014-08-29
CN201410437102.XA CN104156489B (zh) 2014-08-29 2014-08-29 基于司机轨迹进行司机常驻点挖掘的方法
CN201410705608.4A CN104574947A (zh) 2014-11-27 2014-11-27 处理订单的方法和设备
CN201410705608.4 2014-11-27
CN201510020526.0A CN104537502A (zh) 2015-01-15 2015-01-15 处理订单的方法和设备
CN201510020526.0 2015-01-15
CN201510163063.3 2015-04-08
CN201510163063.3A CN104715426B (zh) 2015-04-08 2015-04-08 用于处理订单的方法及设备

Publications (1)

Publication Number Publication Date
WO2016019857A1 true WO2016019857A1 (zh) 2016-02-11

Family

ID=55263155

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/086075 WO2016019857A1 (zh) 2014-08-04 2015-08-04 服务派发系统及方法

Country Status (7)

Country Link
US (1) US20170228683A1 (zh)
EP (1) EP3179420A4 (zh)
KR (2) KR20190000400A (zh)
MY (1) MY188692A (zh)
PH (1) PH12017500192A1 (zh)
SG (2) SG11201700895YA (zh)
WO (1) WO2016019857A1 (zh)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108229879A (zh) * 2017-12-26 2018-06-29 拉扎斯网络科技（上海）有限公司 一种行程时长预估方法、装置及存储介质
CN110322141A (zh) * 2019-06-28 2019-10-11 浙江吉利控股集团有限公司 一种实时单自动接单方法、装置及终端
US10977585B2 (en) 2015-01-29 2021-04-13 Beijing Didi Infinity Technology And Development Co., Ltd. Order allocation system and method
CN113379484A (zh) * 2021-05-31 2021-09-10 五八到家有限公司 订单快速匹配的方法、系统、设备及介质
CN114239278A (zh) * 2021-12-17 2022-03-25 中国科学院生态环境研究中心 土壤重金属累积过程时空模拟模型构建方法
TWI808594B (zh) * 2022-01-03 2023-07-11 郭朝誠 具租借共享交通工具功能的外送服務媒合系統

Families Citing this family (37)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2016035091A1 (en) * 2014-09-03 2016-03-10 Meru Cab Company Private Limited Dynamic forecasting for forward reservation of cab
NZ733863A (en) 2015-01-20 2019-04-26 Beijing Didi Infinity Technology & Dev Co Ltd Systems and methods for providing information for an on-demand service
GB2550523A (en) 2015-02-02 2017-11-22 Beijing Didi Infinity Tech And Dev Co Ltd Methods and systems for order processing
US10635994B2 (en) 2015-12-11 2020-04-28 Lyft, Inc. System for navigating driver to passenger for ride authorized by another user of transportation service
US11508026B2 (en) 2015-12-31 2022-11-22 Lyft, Inc. System for navigating transportation service providers to fulfill transportation requests authorized by an organization
CN106384509A (zh) * 2016-10-08 2017-02-08 大连理工大学 考虑出租车运营状态下的城市道路行程时间分布估算方法
CN108009180B (zh) * 2016-10-28 2021-09-21 哈尔滨工业大学深圳研究生院 一种高质量模式项集挖掘方法、装置及数据处理设备
CN106781468B (zh) * 2016-12-09 2018-06-15 大连理工大学 基于建成环境和低频浮动车数据的路段行程时间估计方法
AU2018290417A1 (en) * 2017-06-28 2019-01-24 Beijing Didi Infinity Technology And Development Co., Ltd. Systems and methods for cheat examination
EP3653015A4 (en) * 2017-07-14 2020-12-02 Telefonaktiebolaget LM Ericsson (PUBL) COMMUNICATION PROCESS AND DEVICE FOR DETERMINING A SCORE RELATING TO A FIRST AGENT MODULE
US20190102723A1 (en) * 2017-10-02 2019-04-04 Servicenow, Inc. Systems for automated profile building, skillset identification, and service ticket routing
CN110800030B (zh) 2017-10-31 2022-09-30 北京嘀嘀无限科技发展有限公司 用于拼车服务的方法和系统
CN107945058A (zh) * 2017-11-08 2018-04-20 北京轻生活时代科技发展有限责任公司 一种智能点餐管理系统
WO2019113976A1 (en) * 2017-12-15 2019-06-20 Beijing Didi Infinity Technology And Development Co., Ltd. Systems and methods for optimizing online on-demand and service
US11475395B2 (en) 2018-01-19 2022-10-18 Walmart Apollo, Llc Systems and methods for combinatorial resource optimization
US20190228352A1 (en) 2018-01-19 2019-07-25 Walmart Apollo, Llc Systems and methods for combinatorial resource optimization
CN110309241B (zh) * 2018-03-15 2021-11-23 北京嘀嘀无限科技发展有限公司 挖掘方法、装置、服务器、计算机设备和可读存储介质
WO2019242286A1 (en) * 2018-06-19 2019-12-26 Beijing Didi Infinity Technology And Development Co., Ltd. Systems and methods for allocating service requests
CN110785749B (zh) * 2018-06-25 2020-08-21 北京嘀嘀无限科技发展有限公司 用于生成宽表的系统和方法
CN109033219B (zh) * 2018-06-29 2022-03-11 北京奇虎科技有限公司 兴趣点poi分类方法和装置
CN108965462B (zh) * 2018-08-06 2019-06-21 长安大学 基于群代理协同交互的共乘出行仿真系统及其实现方法
US11443355B2 (en) 2018-08-23 2022-09-13 FoodMaven, Inc. Method and system for presenting food alternatives to food buyers
CN110875937A (zh) * 2018-08-31 2020-03-10 北京嘀嘀无限科技发展有限公司 一种信息推送方法和系统
CN109460851A (zh) * 2018-09-26 2019-03-12 长沙市到家悠享家政服务有限公司 订单分配方法、装置及电子设备
CN109522449B (zh) * 2018-09-28 2021-11-30 百度在线网络技术（北京）有限公司 搜索方法和装置
US10977441B2 (en) * 2018-10-29 2021-04-13 Amazon Technologies, Inc. Normalizing addresses to facilitate sortation and routing solution using natural language text processing
US11615368B2 (en) * 2018-11-01 2023-03-28 Walmart Apollo, Llc Systems and methods for determining delivery time and route assignments
CN111144676A (zh) * 2018-11-05 2020-05-12 北京嘀嘀无限科技发展有限公司 用车订单分配方法、装置、服务器及计算机可读存储介质
CN111178647B (zh) * 2018-11-09 2023-09-05 中移（杭州）信息技术有限公司 一种推送工单的方法、系统及计算机存储介质
CN111353732A (zh) * 2018-12-21 2020-06-30 北京中交兴路信息科技有限公司 车辆运输模式识别方法、装置、设备和存储介质
JP7131422B2 (ja) * 2019-02-13 2022-09-06 トヨタ自動車株式会社 情報処理装置
US11574377B2 (en) 2019-06-03 2023-02-07 International Business Machines Corporation Intelligent on-demand management of ride sharing in a transportation system
CN110578999B (zh) * 2019-07-12 2021-09-10 广东美的制冷设备有限公司 空调器及其控制方法、装置和计算机可读存储介质
US11074548B2 (en) 2019-12-05 2021-07-27 Coupang Corp. Computer implemented systems and methods for optimization of a product inventory by intelligent distribution of inbound products
US11599846B2 (en) * 2020-07-02 2023-03-07 Kpn Innovations, Llc. Method and system for selection of a path for deliveries
CN112286982B (zh) * 2020-09-25 2024-01-30 长沙市到家悠享网络科技有限公司 数据处理方法、服务端设备以及客户端设备
CN112413859B (zh) * 2020-11-24 2021-11-16 珠海格力电器股份有限公司 空调器控制方法、装置、空调器及存储介质

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102737501A (zh) * 2012-06-12 2012-10-17 中国联合网络通信集团有限公司 出租车载客调度方法和系统以及调度服务器
JP2013076643A (ja) * 2011-09-30 2013-04-25 Hitachi Solutions Ltd 配車タクシー検索システム及びプログラム並びに車載端末
CN103218769A (zh) * 2013-03-19 2013-07-24 王兴健 出租车订单分配方法
CN104157133A (zh) * 2014-08-20 2014-11-19 北京嘀嘀无限科技发展有限公司 基于司机在线活跃情况的运力拉升系统
CN104156868A (zh) * 2014-08-22 2014-11-19 北京嘀嘀无限科技发展有限公司 基于订单价值判断促进订单成交的出租车积分系统
CN104574947A (zh) * 2014-11-27 2015-04-29 北京嘀嘀无限科技发展有限公司 处理订单的方法和设备
CN104715426A (zh) * 2015-04-08 2015-06-17 北京嘀嘀无限科技发展有限公司 用于处理订单的方法及设备

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100397434C (zh) * 2004-08-07 2008-06-25 中华电信股份有限公司 出租车营运安全与派遣监控系统
US20090313077A1 (en) * 2008-06-17 2009-12-17 Wheeler Iv George Y Consumer initiated, service provider direct dispatching system
US20100158202A1 (en) * 2008-12-23 2010-06-24 International Business Machines Corporation Location Based Emergency Services Dispatching
CN101819045A (zh) * 2009-02-27 2010-09-01 联发科技（新加坡）私人有限公司 应用于导航装置的道路选择方法
US20140229258A1 (en) * 2011-03-16 2014-08-14 Malak Seriani Systems and methods enabling transportation service providers to competitively bid in response to customer requests
SG195395A1 (en) * 2012-05-07 2013-12-30 Accenture Global Services Ltd Location-based cognitive and predictive communication system
WO2014074513A1 (en) * 2012-11-06 2014-05-15 Intertrust Technologies Corporation Activity recognition systems and methods
CN103514738B (zh) * 2013-09-18 2015-09-02 福建工程学院 一种出租车叫车拼车顺路匹配方法及其系统

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2013076643A (ja) * 2011-09-30 2013-04-25 Hitachi Solutions Ltd 配車タクシー検索システム及びプログラム並びに車載端末
CN102737501A (zh) * 2012-06-12 2012-10-17 中国联合网络通信集团有限公司 出租车载客调度方法和系统以及调度服务器
CN103218769A (zh) * 2013-03-19 2013-07-24 王兴健 出租车订单分配方法
CN104157133A (zh) * 2014-08-20 2014-11-19 北京嘀嘀无限科技发展有限公司 基于司机在线活跃情况的运力拉升系统
CN104156868A (zh) * 2014-08-22 2014-11-19 北京嘀嘀无限科技发展有限公司 基于订单价值判断促进订单成交的出租车积分系统
CN104574947A (zh) * 2014-11-27 2015-04-29 北京嘀嘀无限科技发展有限公司 处理订单的方法和设备
CN104715426A (zh) * 2015-04-08 2015-06-17 北京嘀嘀无限科技发展有限公司 用于处理订单的方法及设备

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3179420A4 *

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10977585B2 (en) 2015-01-29 2021-04-13 Beijing Didi Infinity Technology And Development Co., Ltd. Order allocation system and method
CN108229879A (zh) * 2017-12-26 2018-06-29 拉扎斯网络科技（上海）有限公司 一种行程时长预估方法、装置及存储介质
CN108229879B (zh) * 2017-12-26 2021-01-05 拉扎斯网络科技（上海）有限公司 一种行程时长预估方法、装置及存储介质
CN110322141A (zh) * 2019-06-28 2019-10-11 浙江吉利控股集团有限公司 一种实时单自动接单方法、装置及终端
CN113379484A (zh) * 2021-05-31 2021-09-10 五八到家有限公司 订单快速匹配的方法、系统、设备及介质
CN113379484B (zh) * 2021-05-31 2023-06-06 五八到家有限公司 订单快速匹配的方法、系统、设备及介质
CN114239278A (zh) * 2021-12-17 2022-03-25 中国科学院生态环境研究中心 土壤重金属累积过程时空模拟模型构建方法
CN114239278B (zh) * 2021-12-17 2022-06-21 中国科学院生态环境研究中心 土壤重金属累积过程时空模拟模型构建方法
TWI808594B (zh) * 2022-01-03 2023-07-11 郭朝誠 具租借共享交通工具功能的外送服務媒合系統

Also Published As

Publication number Publication date
KR20190000400A (ko) 2019-01-02
SG10201901024TA (en) 2019-03-28
US20170228683A1 (en) 2017-08-10
PH12017500192B1 (en) 2017-06-28
MY188692A (en) 2021-12-23
EP3179420A4 (en) 2018-03-14
EP3179420A1 (en) 2017-06-14
SG11201700895YA (en) 2017-03-30
KR20180006871A (ko) 2018-01-19
PH12017500192A1 (en) 2017-06-28

Similar Documents

Publication Publication Date Title
WO2016019857A1 (zh) 服务派发系统及方法
JP6918087B2 (ja) オン・デマンドサービスの情報を提供する方法及びシステム
US11315170B2 (en) Methods and systems for order processing
TWI638320B (zh) 推薦預計到達時間的系統、方法及非暫態電腦可讀儲存媒體
WO2016127918A1 (zh) 一种运力调度方法及系统
WO2016138863A1 (zh) 一种订单配对系统及方法
WO2016119749A1 (zh) 一种订单分配系统及方法
JP2018530022A (ja) 個人別コンテンツを推奨するためのシステム及び方法
US10820141B2 (en) Method and apparatus for presenting privacy-respectful and personalized location-based comments based on passenger context and vehicle proximity to the location
CA2976122A1 (en) Methods and systems for pushing orders
CN111861622A (zh) 一种确定上车点推荐模型的方法和系统
JP6098302B2 (ja) ナビゲーションシステム、ナビゲーション方法、及びナビゲーションプログラム
CN114041129A (zh) 确定上车点名称的系统和方法
JP6956232B2 (ja) 探索システム、探索方法、および探索プログラム
US20220248170A1 (en) Methods and systems for recommending pick-up points
JP6048196B2 (ja) ナビゲーションシステム、ナビゲーション方法、及びナビゲーションプログラム
CN110651266B (zh) 为按需服务提供信息的系统及方法
US20220397408A1 (en) Content Delivery In Real-Time Guided Navigation
Nguyen Deriving trip's modes and trip's purposes from GPS-based travel surveys
JP2014178299A (ja) ナビゲーションシステム、ナビゲーション方法、及びナビゲーションプログラム
Hu et al. Understanding the potential of taxi sharing: The case of Chengdu
BR112017016064B1 (pt) Métodos e sistemas para fornecer informação para um serviço por demanda
CN116797752A (zh) 一种地图渲染方法、装置、电子设备及存储介质
NZ751377B2 (en) Methods and systems for providing information for an on-demand service
NZ751377A (en) Methods and systems for providing information for an on-demand service

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15829451

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 12017500192

Country of ref document: PH

REEP Request for entry into the european phase

Ref document number: 2015829451

Country of ref document: EP

WWE Wipo information: entry into national phase

Ref document number: 2015829451

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 20177003867

Country of ref document: KR

Kind code of ref document: A