CN110458429A - A kind of intelligent task distribution and personal scheduling method, system for geographical site - Google Patents
A kind of intelligent task distribution and personal scheduling method, system for geographical site Download PDFInfo
- Publication number
- CN110458429A CN110458429A CN201910689985.6A CN201910689985A CN110458429A CN 110458429 A CN110458429 A CN 110458429A CN 201910689985 A CN201910689985 A CN 201910689985A CN 110458429 A CN110458429 A CN 110458429A
- Authority
- CN
- China
- Prior art keywords
- visit
- site
- business personnel
- task
- feature
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Educational Administration (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of for the Intelligent visiting task distribution of the site with certain geographic location feature and personal scheduling method, system, the following steps are included: from the basic data and real time data of enterprise information system decision-making module, feature is extracted respectively to physics commercial network and business personnel, the historical data for visiting site to business personnel models;The connection between business personnel and site is indicated with bipartite graph to form call plan network;According to business personnel's local environment information, optimal visit route is planned;Using intensified learning method, the preset award or punishment generated after a certain movement, the income that assessment allocation strategy obtains are executed according to business personnel, and feeds back to deep neural network model, and renewal learning parameter, determines OPTIMAL TASK matching strategy repeatedly;It during practical visit, is calculated using intensified learning method, by task execution policy optimization method and visit route optimization method while continuous iteration, the comprehensive visit mode for obtaining global optimum.
Description
Technical field
The present invention relates to enterprise information system research field, in particular to a kind of intelligent task for geographical site is distributed
With personal scheduling method, system.
Background technique
Information system is the brain of enterprise operation.Gradually along with scope of the enterprise expansion, IT application in enterprise process
Deeply, management depth, data and information content constantly increase.Therefore entirely with labor management system information and business decision, enterprise is raw
High efficiency is much not achieved in production.There is many simple and high frequency empirical decision problems, letter especially in enterprise's day-to-day operations
There is also a large amount of empirical decision-making modules in breath system, such issues that often expend a large amount of manpowers and be not required to excessive mental, increase
The human cost of enterprise.
The typical call plan such as towards commercial network under line, the decision for being frequently present of task distribution and personal scheduling are asked
Topic, due to there is uncertain and complexity in reality, original allocation scheme lacks scientific basis, while to environment and friendship
Understanding and considerate condition is insensitive, may become unreasonable, and at this moment the general solution of enterprise is to take manually to plan again, then be passed to enterprise
In industry information system, and notify business personnel that task changes, this sequence of operations not only influences the working efficiency of business personnel,
It returns client and leaves bad impression.
With the continuous development of artificial intelligence technology, the methods of deep learning and intensified learning obtain artificial intelligence newly
It breaks through, artificial intelligence technology is applied into enterprise information system, carry out empirical decision instead of people, improve the flexibility in decision of enterprise
Change.
Based on background above, a kind of achievable enterprise information system intelligence response is studied, it can the distribution of real-time perfoming task
There is important practical value with the method for personal scheduling.
Summary of the invention
The shortcomings that it is a primary object of the present invention to overcome existing method and deficiency, provide a kind of intelligence for geographical site
The distribution of energy task and personal scheduling method, system, the user's big data accumulated using enterprise are especially needing manually
In the module for carrying out empirical operational decision making, extract service logic rule, input and output data to this part of module into
Row modeling, training deep neural network overcome the cost of decision making of manual decision to replace manual decision using neural network
The problems such as height, low efficiency.
The purpose of the present invention is realized by the following technical solution: a kind of for the intelligent task distribution of geographical site and people
Member's dispatching method, comprising steps of
(1) from the basic data and real time data of enterprise information system decision-making module, feature is extracted, feature includes wait visit
Visit the feature, the feature of business personnel, feature, the business personnel's surrounding enviroment feature of visiting task etc. of site.
(2) use bipartite graph by the commercial network (hereinafter referred to as site) with geographic location feature of business personnel and visit
Between relationship form call plan network, the real time data for visiting site to historical record and business personnel models, and obtains depth
Intensified learning model;
(3) according to the preset award or punishment that generate after a certain movement is executed, the income that allocation strategy obtains is assessed, and
Deeply learning model is fed back to, iterate renewal learning parameter, final to determine OPTIMAL TASK implementation strategy;
(4) according to business personnel's local environment information, optimal visit route is planned;
(5) it during practical visit, is calculated using intensified learning method, the task execution strategy of step (3) is excellent
Continuous iteration, synthesis obtain the visit mode of global optimum to the visit route optimization method of change method and step (4) simultaneously.
Preferably, after the basic data for obtaining decision-making module, data prediction is first carried out to it, pretreatment includes: to check
It whether there is dirty data in initial data, dirty data generally refers to undesirable, and cannot directly carry out corresponding analysis
Data;It fills in the value of missing, smooth noise data, identification or deletes outlier and solve inconsistency to clear up data;Pass through
The modes such as smooth aggregation, standardization convert the data into the form suitable for neural metwork training.
Preferably, in step (1), the basic data of decision-making module refers to the historical data of service database, including wait visit
Site information, business personnel's information etc. are visited, real time data refers to visit mission bit stream, real-time traffic condition, business personnel geographical location
Deng.
Preferably, in step (1), the feature of the site to be visited includes: that visit site type and site are believed substantially
Breath, the visit site type include existing site, potential site, doubtful false site etc..The site essential information includes ground
Manage position, manage category, mode of supplly, product inventory etc..
The feature of the business personnel includes: professional ability, history visit record, known site and range etc..
The feature of the visit task include: visit sequence and route, it is contemplated that order target, between site and business personnel
Interaction scenario, i.e., the site whether by business personnel visit, visit task difficulty, business personnel is visiting or clear operation.
For whole process is visited in site, business personnel, which visits a site, need to undergo three processes, that is, goes to site to be visited, visits
It visits site and leaves site.
Business personnel's surrounding enviroment feature includes: time and the spatiality of business personnel, that is, be presently in geographical location,
Traffic condition, visit time etc..
Preferably, in step (1), when extracting the feature wait visit site, to each mesh point sets sequencing.
Preferably, in step (2), the input of deeply learning model is state set S and behavior aggregate A, the spy of business personnel
Seek peace business personnel surrounding enviroment character representation state set S, the feature of site to be visited and the character representation behavior aggregate of visit task
A。
Preferably, in step (2), during the deeply learning model of foundation, business is indicated with bipartite graph
Matching relationship between member and visit site, in the bipartite graph, using business personnel and site as node, business personnel and site
It is initially full connection status, in the interactive process with environment, as the weight on connection side constantly alternates, meanwhile, base area
Reason positional relationship establishes partial ordering relation to site set, obtains best fit strategy.
Further, in step (2), during carrying out visit site and matched business personnel, some realities are considered as
Border situation, such as existing site visit are visited better than potential site, (Ti,Tj) indicate that site i must be visited before the j of site, one
The unlatching of visit task must be after the completion of a upper call plan, i.e. arrival site j starts the time f visitedr(j) be less than from
Open the time f of site iL(i)。
Preferably, in step (3), it is assumed that visit acts a, can make business personnel that shape occur automatically after completing visit movement
The transfer of state, i.e., from stIt is transformed into st+1, reward can be brought after being completed at the same time visit movement, definition reward r is visit site
Total target performance for executing the time and visiting site;The information is fed back into deeply learning model;According to anti-
Feedforward information calculate for indicating expection accumulated earnings under certain state desired value of the movements function Q (s, a), the value letter
Number indicates the weight that side is connected between business personnel and site, i.e. learning parameter, constantly change cost function, solves Optimum Matching,
And then determine OPTIMAL TASK implementation strategy.
Preferably, in step (4), after determining OPTIMAL TASK implementation strategy, visit route planning is carried out, i.e., according to real-time road
The ambient environmental factors such as condition, distance plan optimal site visit sequence and path.
Preferably, in step (5), during practical visit, feelings are completed in the position and call plan for assessing business personnel in real time
Condition analyses whether the risk for being unable to complete daily regulation visit site, reassigns in time;Daily existing site is preferentially visited, then
Implement the position of visit situation and business personnel according to business personnel, distributes new shop leaved for development.
For the intelligent task distribution of geographical site and personal scheduling system, comprising:
Characteristic extracting module extracts feature from the basic data and real time data of enterprise information system decision-making module, packet
The feature of feature, business personnel containing site to be visited, feature, the business personnel's surrounding enviroment feature for visiting task;
Deeply learning model constructs module, for forming visit meter to the relationship between business personnel and visit site
Two subnetworks drawn, while partial ordering relation is established to site set according to geographical positional relationship, historical record and business personnel are visitd
The real time data modeling for visiting site, obtains deep neural network model;
Task matching strategy optimization module, for according to business personnel execute the effect generated after a certain movement determine it is preset
Award or punishment, the income that assessment allocation strategy obtains, and deeply learning model is fed back to, renewal learning parameter determines
OPTIMAL TASK implementation strategy
Route optimization module is visited, for planning optimal visit route according to business personnel's local environment information;
Visit mode determining module holds the task in task matching strategy optimization module during practical visit
Visit route optimization method while continuous iteration in row policy optimization method and task matching strategy optimization module, it is comprehensive
Obtain the visit mode of global optimum.
Compared with the prior art, the invention has the following advantages and beneficial effects:
(1) present invention is directed to original manual decision business personnel Task Allocation Problem, needs people to the original of enterprise information system
The data that output and input in the module of work decision carry out analysis modeling, train deeply learning model using historical data,
Realize enterprise information system intelligence response, thus make neural network replace manual decision, be enterprise's Real-Time Scheduling business personnel and
Distribution task provides decision scheme.
(2) artificial intelligence technology is applied into enterprise information system, carries out empirical decision instead of people, reduces enterprise
Manpower cost of decision making optimizes the flexibility in decision of enterprise.
(3) make neural network that manual decision be replaced to realize enterprise information system intelligence response, it can Real-Time Scheduling business personnel
And distribution task, the time of business decision task matching plan is shortened, the working efficiency of business personnel is promoted.
(4) it is combined in real time using visit task matching optimization and path optimization's algorithm, and ambient environmental factors is considered
Inside, visit task matching precision is improved, and promotes visit efficiency.
(5) priority ordered is carried out to task, for the reserved visit blank time of business personnel, priority is lower non-need to promptly to be visitd
The site of visit, according to the actual conditions that business personnel visits, the distribution of real-time perfoming task.
Detailed description of the invention
Fig. 1 is the business personnel and site matching figure of the method for the present invention.
Fig. 2 is the process when present invention carries out deeply learning training.
Fig. 3 is the flow diagram of the method for the present invention.
Specific embodiment
Present invention will now be described in further detail with reference to the embodiments and the accompanying drawings, but embodiments of the present invention are unlimited
In this.
Embodiment 1
Selling the task distribution of site and dispatching method, step under industry business personnel visit line fastly the present embodiment provides one kind is
Information system related data is extracted first, extracts feature, and then is indicated business personnel with bipartite graph and visited the pass between site
System forms call plan network.The matching of daily visit site task is carried out by history service data in enterprise information system,
In conjunction with business personnel's current state in network, historical data and real time data are appointed from visit using deeply learning method
Be engaged in matching optimization and the two levels of business personnel path optimization continuous iteration simultaneously, the comprehensive visit mode for obtaining global optimum and
Visit path.Each step is specifically described with reference to the accompanying drawing.
One, from the data and real time data of enterprise information system decision-making module, feature is extracted.
The module of enterprise information system is divided, the basic data of empirical manual decision's module is handled,
The partial data includes site information to be visited, business personnel's information etc..Real time data refers to visit mission bit stream, real-time traffic shape
Condition, business personnel geographical location etc..
Historical data and real time data to service database are analyzed and are handled with deep learning method, extract to
Visit the feature, the feature of business personnel, feature, the business personnel's surrounding enviroment feature for visiting task of site.The spy of site to be visited
Sign includes: visit site type and site essential information, and the visit site type includes existing site, potential net to be developed
Point, doubtful false site, the site information include geographical location, operation category, mode of supplly, product inventory etc..Carrying out industry
When business person matches with visit site, the daily visit task of existing site is paid the utmost attention to, potential site visit is according to daily visit
The performance of task, in real time arranged to business personnel.
The feature of business personnel includes: professional ability, history visit record, known site and range etc..Visit task
Feature includes: the interaction scenario between site and business personnel, i.e., the site whether by business personnel visit, visit task difficulty,
Expected visit effect etc..For whole process is visited in site, business personnel, which visits a site, need to undergo three processes, i.e., before
Toward site to be visited, visit site and leave site.The surrounding enviroment of business personnel include: time and the space shape of business personnel
State is presently in geographical location, traffic condition, visit time etc..
Two, it is modeled, is used using the real time data that deeply learning method visits site to historical record and business personnel
Bipartite graph indicates business personnel and visits the matching relationship between site.
To deeply method study training process in state set S and behavior aggregate A be defined, the feature of business personnel and
Business personnel's surrounding enviroment indicate state set S, the feature of site to be visited and the character representation behavior aggregate A of visit task.
In the model of foundation, with bipartite graph come the matching relationship between abstract service person and site, by business personnel and net
Point is used as node, and the initial link of business personnel and site is full connection status, by the interactive process with environment, with connection
Weight on side constantly alternates, and obtains best fit strategy, such as Fig. 1.During carrying out site and matched business personnel, take an examination
Consider some actual conditions, such as existing site visit is visited better than potential site, (Ti,Tj) indicate that site i must be visited before j.
The unlatching of one visit task must be after the completion of a upper call plan, i.e. arrival site j starts the time f visitedr(j) small
In the time f for leaving site iL(i)。
In the model of foundation, partial ordering relation is established between the geographical location relationship site.
Three, the generation after acting a according to execution is awarded or punishment, the income that assessment allocation strategy obtains, and feeds back to strong
Change learning model, renewal learning parameter determines OPTIMAL TASK implementation strategy by way of constantly feeding back update.
Movement a is the transfer that the movement for completing to visit can make business personnel's generating state automatically, i.e., from stIt is transformed into st+1, together
When execution after can bring reward, definition reward r be the total execution time for visiting site and the feedback for visiting site.
(s a) indicates the expectation of the expection accumulated earnings under certain state to value of the movements function Q in intensified learning, corresponding
It is always executing the time and is visiting the feedback of site, cost function is the power for indicating to connect side between business personnel and site
Weight, constantly change cost function solve Optimum Matching, into final distribution link.
Four, plan that optimal visit route carries out visit route planning, i.e., according to reality after determining daily visit matching strategy
The ambient environmental factors such as Shi Lukuang, distance plan that smooth and path is visited in optimal site.
Five, according to visit actual conditions, visit task is automatically adjusted in conjunction with Task Assigned Policy and paths planning method, and
New shop leaved for development is distributed during visit.
Automatic adjustment visit task is position and the call plan performance of real-time assessment business personnel, analyse whether there is or not
Method completes the risk of daily regulation visit site, reassigns in time.Site to be allocated refers to that daily existing site is preferentially visited, root
Implement the position of visit situation and business personnel according to business personnel, new shop leaved for development of reallocating.
Embodiment 2
A kind of intelligent task distribution and personal scheduling system for geographical site, comprising:
Characteristic extracting module, it is special for extracting from the basic data and real time data of enterprise information system decision-making module
Sign, feature include feature, the feature of business personnel, feature, the business personnel's surrounding enviroment feature of visiting task of site to be visited etc.;
Deeply learning model constructs module, for using bipartite graph for business personnel and visiting special with geographical location
Relationship between the commercial network of sign forms call plan network, and the real time data for visiting site to historical record and business personnel is built
Mould obtains deeply learning model;
Task matching strategy optimization module, preset award or punishment for being generated according to business personnel's implementation effect, is commented
Estimate the income of allocation strategy acquisition, and feed back to deeply learning model, renewal learning parameter determines that OPTIMAL TASK executes plan
Slightly;
Route optimization module is visited, for planning optimal visit route according to business personnel's local environment information;
Visit mode determining module, for being calculated using intensified learning method, by task during actually visit
The visit route optimization in task matching strategy optimization method and visit route optimization module in implementation strategy optimization module
Continuous iteration, synthesis obtain the visit mode of global optimum to method simultaneously.
It is apparent to those skilled in the art that for convenience of description and succinctly, foregoing description is
The specific work process of system, module, can refer to corresponding processes in the foregoing method embodiment, details are not described herein.
Those of ordinary skill in the art may be aware that list described in conjunction with the examples disclosed in the embodiments of the present disclosure
Member and algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, in order to clearly demonstrate hardware
With the interchangeability of software, each exemplary composition and step are generally described according to function in the above description.This
A little functions are implemented in hardware or software specific application and design constraint depending on technical solution actually.Profession
Technical staff can use different methods to achieve the described function each specific application, but this realization is not answered
Think beyond the scope of this invention.
In embodiment provided by the present invention, it should be understood that disclosed system, module and method can pass through
Other modes are realized.For example, the apparatus embodiments described above are merely exemplary, for example, the division of the module,
Only logical function partition, there may be another division manner in actual implementation, can also be by module with the same function
A module is assembled, such as multiple module or components can be combined or can be integrated into another system or some features
It can ignore, or not execute.In addition, shown or discussed mutual coupling, direct-coupling or communication connection can be
Through some interfaces, the indirect coupling or communication connection of device or unit, is also possible to electricity, and mechanical or other forms connect
It connects.
The module as illustrated by the separation member may or may not be physically separated, aobvious as module
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.Some or all of the modules therein can be selected to realize the embodiment of the present invention according to the actual needs
Purpose.
It, can also be in addition, each functional module in each embodiment of the present invention can integrate in one processing unit
It is that modules physically exist alone, is also possible to two or more modules and is integrated in a module.It is above-mentioned integrated
Unit both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated module is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in one storage medium.Based on this understanding, technical solution of the present invention is substantially in other words to existing
The all or part of part or the technical solution that technology contributes can be embodied in the form of software products, should
Computer software product is stored in a storage medium, including some instructions are used so that a computer equipment (can be
Personal computer, server or network equipment etc.) execute all or part of step of each embodiment the method for the present invention
Suddenly.And storage medium above-mentioned include: USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), magnetic disk or
The various media that can store program code such as person's CD.
The above embodiment is a preferred embodiment of the present invention, but embodiments of the present invention are not by above-described embodiment
Limitation, other any changes, modifications, substitutions, combinations, simplifications made without departing from the spirit and principles of the present invention,
It should be equivalent substitute mode, be included within the scope of the present invention.
Claims (8)
1. a kind of for the intelligent task distribution of geographical site and personal scheduling method, which is characterized in that step are as follows:
(1) feature is extracted from the basic data and real time data of enterprise information system decision-making module, feature includes net to be visited
The feature of point, the feature of business personnel, feature, the business personnel's surrounding enviroment feature for visiting task;
(2) to the relationship between business personnel and visit site, two subnetworks of call plan are formed, while closing according to geographical location
System establishes partial ordering relation to site set, and the real time data for visiting site to historical record and business personnel models, and obtains depth mind
Through network model;
(3) effect generated after a certain movement is executed according to business personnel and determines preset award or punishment, assessment allocation strategy obtains
The income taken, and deeply learning model is fed back to, renewal learning parameter determines OPTIMAL TASK implementation strategy;
(4) according to business personnel's local environment information, optimal visit route is planned;
(5) during practical visit, the visit route of the task execution policy optimization method of step (3) and step (4) is excellent
Continuous iteration, synthesis obtain the visit mode of global optimum to change method simultaneously.
2. task distribution according to claim 1 and personal scheduling method, which is characterized in that described wait visit in step (1)
The feature for visiting site includes: visit site type and site essential information, and the visit site type includes existing site, dives
In site to be developed, doubtful false site, the site information includes operator's basic information, geographical location, manages category, the supply of material
Mode, product inventory;The feature of the business personnel includes: basic information, professional ability, history visit record, known site
With range, known product classification;The feature of the visit task includes: the interaction scenario between site and business personnel, i.e., should
Whether site is by business personnel's visit, difficulty, the target of visit of visit task;The surrounding enviroment of the business personnel include:
The time of business personnel and spatiality, i.e., locating geographical location, traffic condition, visit time.
3. task distribution according to claim 1 and personal scheduling method, which is characterized in that in step (2), neural network
The input of model is state set S and behavior aggregate A, and the feature and business personnel's surrounding enviroment of business personnel indicates state set S, net to be visited
The feature of point and the character representation behavior aggregate A of visit task.
4. task distribution according to claim 1 and personal scheduling method, which is characterized in that in step (2), in foundation
During deep neural network model, indicated with bipartite graph business personnel and visit site between matching relationship, this two
In component, using business personnel and site as node, business personnel and site are initially full connection status, are interacted with environment
Cheng Zhong, as the weight on connection side constantly alternates, meanwhile, partial ordering relation is established to site set according to geographical positional relationship,
Obtain best fit strategy.
5. task distribution according to claim 1 and personal scheduling method, which is characterized in that in step (3), complete this and visit
Reward can be brought after visit movement a, reward r is defined as total execution time of visit site and the effect of visit site, is completed at the same time
The transfer that can make business personnel's generating state after visit movement automatically, i.e., from stIt is transformed into st+1;The feedback informations such as effect are fed back
To model, for calculate the expection accumulated earnings under certain state desired value of the movements function Q (s, a), the cost function
It indicates the weight for connecting side between business personnel and site, i.e. learning parameter, constantly change cost function, solves Optimum Matching, into
And determine OPTIMAL TASK implementation strategy.
6. task distribution according to claim 1 and personal scheduling method, which is characterized in that in step (4), determine optimal
After task execution strategy, visit route planning is carried out, i.e., plans optimal site visit sequence and path according to ambient environmental factors,
Ambient environmental factors include real-time road, distance.
7. task distribution according to claim 1 and personal scheduling method, which is characterized in that practical to visit in step (5)
In the process, position and the call plan performance for assessing business personnel in real time have analysed whether to be unable to complete daily regulation visit
The risk of site, is reassigned in time;Daily existing site is preferentially visited, and implements the position of visit situation and business personnel further according to business personnel
It sets, distributes new shop leaved for development.
8. a kind of for the intelligent task distribution of geographical site and personal scheduling system, which is characterized in that
Characteristic extracting module extracts feature from the basic data and real time data of enterprise information system decision-making module, comprising to
Visit the feature, the feature of business personnel, feature, the business personnel's surrounding enviroment feature for visiting task of site;
Deeply learning model constructs module, for forming call plan to the relationship between business personnel and visit site
Two subnetworks, while partial ordering relation is established to site set according to geographical positional relationship, net is visited to historical record and business personnel
The real time data modeling of point, obtains deep neural network model;
Task matching strategy optimization module determines preset award for executing the effect generated after a certain movement according to business personnel
Or punishment, the income that assessment allocation strategy obtains, and deeply learning model is fed back to, renewal learning parameter determines optimal
Task execution strategy
Route optimization module is visited, for planning optimal visit route according to business personnel's local environment information;
Visit mode determining module, during practical visit, by the task execution plan in task matching strategy optimization module
Continuous iteration, synthesis obtain complete visit route optimization method slightly in optimization method and task matching strategy optimization module simultaneously
The optimal visit mode of office.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910689985.6A CN110458429A (en) | 2019-07-29 | 2019-07-29 | A kind of intelligent task distribution and personal scheduling method, system for geographical site |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910689985.6A CN110458429A (en) | 2019-07-29 | 2019-07-29 | A kind of intelligent task distribution and personal scheduling method, system for geographical site |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110458429A true CN110458429A (en) | 2019-11-15 |
Family
ID=68483939
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910689985.6A Pending CN110458429A (en) | 2019-07-29 | 2019-07-29 | A kind of intelligent task distribution and personal scheduling method, system for geographical site |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110458429A (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111311125A (en) * | 2020-03-25 | 2020-06-19 | 中国建设银行股份有限公司 | Method and device for selecting resource linkage scheme among bank outlets based on genetic algorithm |
CN111553600A (en) * | 2020-04-29 | 2020-08-18 | 广元量知汇科技有限公司 | Big data-based smart city business distribution system |
CN112529325A (en) * | 2020-12-18 | 2021-03-19 | 平安银行股份有限公司 | Network point service distribution method, device, computer equipment and storage medium |
CN112884388A (en) * | 2019-11-29 | 2021-06-01 | 顺丰科技有限公司 | Training method, device and equipment for management strategy generation model |
CN113203427A (en) * | 2021-05-14 | 2021-08-03 | 杉数科技(北京)有限公司 | Path planning method and device and server |
CN113204916A (en) * | 2021-04-15 | 2021-08-03 | 特斯联科技集团有限公司 | Intelligent decision method and system based on reinforcement learning |
CN113657844A (en) * | 2021-06-15 | 2021-11-16 | 中国人民解放军63920部队 | Method and device for determining task processing flow |
CN113671827A (en) * | 2021-07-20 | 2021-11-19 | 大连海事大学 | Dynamic bipartite graph distribution length decision method based on recurrent neural network and reinforcement learning |
WO2022042262A1 (en) * | 2020-08-31 | 2022-03-03 | 中兴通讯股份有限公司 | Otn network resource optimization method and apparatus, computer device, and storage medium |
CN114140033A (en) * | 2022-01-29 | 2022-03-04 | 北京新唐思创教育科技有限公司 | Service personnel allocation method and device, electronic equipment and storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170116522A1 (en) * | 2015-10-05 | 2017-04-27 | Telekom Malaysia Berhad | Method For Task Scheduling And Resources Allocation And System Thereof |
CN106970615A (en) * | 2017-03-21 | 2017-07-21 | 西北工业大学 | A kind of real-time online paths planning method of deeply study |
CN108596335A (en) * | 2018-04-20 | 2018-09-28 | 浙江大学 | A kind of adaptive crowdsourcing method based on deeply study |
CN109697524A (en) * | 2017-10-23 | 2019-04-30 | 顺丰科技有限公司 | Optimization receives and sends part task and resource matched method, system and equipment |
-
2019
- 2019-07-29 CN CN201910689985.6A patent/CN110458429A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170116522A1 (en) * | 2015-10-05 | 2017-04-27 | Telekom Malaysia Berhad | Method For Task Scheduling And Resources Allocation And System Thereof |
CN106970615A (en) * | 2017-03-21 | 2017-07-21 | 西北工业大学 | A kind of real-time online paths planning method of deeply study |
CN109697524A (en) * | 2017-10-23 | 2019-04-30 | 顺丰科技有限公司 | Optimization receives and sends part task and resource matched method, system and equipment |
CN108596335A (en) * | 2018-04-20 | 2018-09-28 | 浙江大学 | A kind of adaptive crowdsourcing method based on deeply study |
Cited By (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112884388A (en) * | 2019-11-29 | 2021-06-01 | 顺丰科技有限公司 | Training method, device and equipment for management strategy generation model |
CN112884388B (en) * | 2019-11-29 | 2023-06-09 | 顺丰科技有限公司 | Training method, device and equipment for management strategy generation model |
CN111311125B (en) * | 2020-03-25 | 2022-06-10 | 中国建设银行股份有限公司 | Method and device for selecting resource linkage scheme among bank outlets based on genetic algorithm |
CN111311125A (en) * | 2020-03-25 | 2020-06-19 | 中国建设银行股份有限公司 | Method and device for selecting resource linkage scheme among bank outlets based on genetic algorithm |
CN111553600A (en) * | 2020-04-29 | 2020-08-18 | 广元量知汇科技有限公司 | Big data-based smart city business distribution system |
WO2022042262A1 (en) * | 2020-08-31 | 2022-03-03 | 中兴通讯股份有限公司 | Otn network resource optimization method and apparatus, computer device, and storage medium |
JP7436747B2 (en) | 2020-08-31 | 2024-02-22 | 中興通訊股▲ふん▼有限公司 | OTN network resource optimization method and apparatus, computing device and storage medium |
CN112529325A (en) * | 2020-12-18 | 2021-03-19 | 平安银行股份有限公司 | Network point service distribution method, device, computer equipment and storage medium |
CN112529325B (en) * | 2020-12-18 | 2024-03-26 | 平安银行股份有限公司 | Network point service distribution method, device, computer equipment and storage medium |
CN113204916B (en) * | 2021-04-15 | 2021-11-19 | 特斯联科技集团有限公司 | Intelligent decision method and system based on reinforcement learning |
CN113204916A (en) * | 2021-04-15 | 2021-08-03 | 特斯联科技集团有限公司 | Intelligent decision method and system based on reinforcement learning |
CN113203427A (en) * | 2021-05-14 | 2021-08-03 | 杉数科技(北京)有限公司 | Path planning method and device and server |
CN113657844A (en) * | 2021-06-15 | 2021-11-16 | 中国人民解放军63920部队 | Method and device for determining task processing flow |
CN113657844B (en) * | 2021-06-15 | 2024-04-05 | 中国人民解放军63920部队 | Task processing flow determining method and device |
CN113671827A (en) * | 2021-07-20 | 2021-11-19 | 大连海事大学 | Dynamic bipartite graph distribution length decision method based on recurrent neural network and reinforcement learning |
CN113671827B (en) * | 2021-07-20 | 2023-06-27 | 大连海事大学 | Dynamic bipartite graph allocation length decision method based on cyclic neural network and reinforcement learning |
CN114140033A (en) * | 2022-01-29 | 2022-03-04 | 北京新唐思创教育科技有限公司 | Service personnel allocation method and device, electronic equipment and storage medium |
CN114140033B (en) * | 2022-01-29 | 2022-04-12 | 北京新唐思创教育科技有限公司 | Service personnel allocation method and device, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110458429A (en) | A kind of intelligent task distribution and personal scheduling method, system for geographical site | |
Kahraman et al. | Intelligence decision systems in enterprise information management | |
DeLaurentis et al. | Taxonomy to guide systems-of-systems decision-making in air transportation problems | |
Sahay et al. | Implementation of GIS in India: Organizational issues and implications | |
CN106155791A (en) | A kind of workflow task dispatching method under distributed environment | |
Giordano et al. | Fuzzy cognitive maps for issue identification in a water resources conflict resolution system | |
Sturtevant et al. | A toolkit modeling approach for sustainable forest management planning: achieving balance between science and local needs | |
CN111191088B (en) | Method, system and readable medium for analyzing cross-boundary service demand | |
Shen et al. | A user pre‐occupancy evaluation method for facilitating the designer‐client communication | |
Feudo | How to build an alternative to sprawl and auto-centric development model through a TOD scenario for the North-Pas-de-Calais region? Lessons from an integrated transportation-land use modelling | |
Calabrese | Communication and co‐operation in product development: a case study of a European car producer | |
CN108876093A (en) | A kind of many many wound design objective method for pushing created under platform | |
Kannan et al. | Competitive performance improvement by operational budget allocation using ANFIS and fuzzy quality function deployment: a case study | |
CN111597176A (en) | Teaching simulation training method and system for delaying supply chain generation | |
Dalal et al. | Simulation modeling at union pacific railroad | |
Dawood | An integrated bidding management expert system for the make-to-order precast industry | |
Derigs et al. | Supporting strategic crew management at passenger railways—model, method and system | |
Tesar et al. | Marketing management in geographically remote industrial clusters: implications for business-to-consumer marketing | |
Lardinois | Simulation, gaming and training in a competitive, multimodal, multicompany, intercity passenger-transportation environment | |
Singh et al. | A simulation-based participatory modelling framework for stakeholder involvement in urban logistics | |
Flamm et al. | Multidisciplinary modeling and GIS for landscape management | |
CN108965462B (en) | Multiply trip analogue system and its implementation altogether based on group agent cooperation interaction | |
Lai et al. | An artificial intelligence approach to course timetabling | |
Escalfoni et al. | Coral framework-a big social data approach to boost startup ecosystem | |
Kozhevnikov et al. | Multi-agent approach for smart resilient city |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191115 |