WO2020107894A1 - 一种驾驶行为评分方法、装置及计算机可读存储介质 - Google Patents

一种驾驶行为评分方法、装置及计算机可读存储介质 Download PDF

Info

Publication number
WO2020107894A1
WO2020107894A1 PCT/CN2019/095164 CN2019095164W WO2020107894A1 WO 2020107894 A1 WO2020107894 A1 WO 2020107894A1 CN 2019095164 W CN2019095164 W CN 2019095164W WO 2020107894 A1 WO2020107894 A1 WO 2020107894A1
Authority
WO
WIPO (PCT)
Prior art keywords
sample
driving
data
stage
insurance policy
Prior art date
Application number
PCT/CN2019/095164
Other languages
English (en)
French (fr)
Inventor
陆璐
梅鵾
谢畅
钱浩然
王恒
孙谷飞
Original Assignee
众安信息技术服务有限公司
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
Application filed by 众安信息技术服务有限公司 filed Critical 众安信息技术服务有限公司
Priority to JP2019554753A priority Critical patent/JP6918137B2/ja
Priority to SG11202005593WA priority patent/SG11202005593WA/en
Publication of WO2020107894A1 publication Critical patent/WO2020107894A1/zh

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
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/08—Insurance
    • 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
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis

Definitions

  • the present disclosure relates to the technical field of vehicle driving behavior analysis, and in particular, to a driving behavior scoring method, device, and computer-readable storage medium.
  • UBI User Based Insurance
  • vehicle insurance based on driving behavior and vehicle usage data pricing. It is a new type of vehicle insurance that appears in the context of liberalization of vehicle insurance rates. Insurance companies collect driving behavior data and vehicle operation data, and after data analysis and processing, assess the driving behavior risk level of drivers. For car owners with good driving behavior, the insurance company can give more discounts, and for car owners with bad driving behavior, the insurance company can increase the premium accordingly. As a result, it helps car owners to cultivate good driving behavior, reduce accidents, reduce risks and claims, thereby improving traffic conditions and reducing social costs.
  • the method of supervised learning in machine learning is usually used to obtain the relationship between driving behavior and driving risk.
  • the training sample needs to contain both driving behavior and vehicle usage data as well as insurance claims data.
  • Collect driving behavior and vehicle usage data usually through on-board OBD (On-Board Diagnostic) or smart phone, etc.
  • OBD On-Board Diagnostic
  • smart phone etc.
  • the vehicle driving data is obtained by sensors in the device, if the device is equipped with a camera, the camera can also obtain driving video Thus, further data reflecting driving habits can be obtained.
  • the risk rate or the compensation rate is used as the indicator of driving risk.
  • the driving data is used as the characteristics of the supervision algorithm, and the risk rate or the compensation rate is used as the supervision signal in the supervision algorithm. Claims relationship. For the target sample that needs to judge the driving behavior risk, that is, the driver who only has driving data, the inferred relationship obtained by the supervision algorithm obtains the driving behavior risk.
  • An object of the embodiments of the present disclosure is to provide a driving behavior scoring method, device, and computer-readable storage medium, and accurately predict driving behavior scores for target drivers by using different driving scoring methods in different stages of data accumulation.
  • the specific technical solutions of the embodiments of the present disclosure are as follows:
  • an embodiment of the present disclosure provides a driving behavior scoring method, the method including:
  • sample data includes sample driver data and sample insurance policy data of the sample driver
  • the accumulated sample data is used to score the driving behavior of the current driver.
  • the sample driving data in the accumulated sample data is used to score the driving behavior of the current driver.
  • the driving scoring method corresponding to the second stage is used to score the driving behavior of the current driver using the sample driving data and the sample insurance policy data in the accumulated sample data.
  • the first stage includes a third stage where the accumulated amount of the sample insurance policy data is below the first threshold and the accumulated amount of the sample driving data is below the second threshold,
  • the scoring of the current driver's driving behavior includes:
  • a rating of the current driver's driving characteristics is obtained as the rating of the current driver's driving behavior.
  • the first stage includes a fourth stage where the accumulation amount of the sample insurance policy data is below the first threshold and the accumulation amount of the sample driving data is above the second threshold,
  • the scoring of the current driver's driving behavior includes:
  • a cumulative distribution function of the driving characteristics of the current driver is calculated as the driving behavior score of the current driver.
  • the second stage includes the first accumulation of the sample insurance policy data above the first threshold, and the accumulation of the sample insurance policy data is less than the accumulation of the sample driving data Five stages,
  • scoring the current driver's driving behavior includes:
  • the risk level cluster corresponding to the current driver is determined, and the determined driving score of the risk level cluster is used as the current driver's driving score.
  • the second stage includes the accumulation of the sample insurance policy data above the first threshold, and the accumulation of the sample insurance policy data is equal to the third of the accumulation of the sample driving data Six stages,
  • the scoring of the current driver's driving behavior includes:
  • the driving behavior scoring model is used to score the driving behavior of the current driver.
  • the obtaining driving scores of a plurality of sample drivers corresponding to the policy data according to the sample insurance policy data includes:
  • a driving score having a mapping relationship with the insurance claim value is determined as the driving score of the sample driver.
  • an embodiment of the present disclosure provides a driving behavior scoring device, the device including:
  • the accumulation module is configured to accumulate sample data, the sample data includes sample driving data of the sample driver and sample insurance policy data;
  • a determination module configured to determine the current accumulation stage of the sample data according to the accumulation amount of the sample driving data and the sample insurance policy data;
  • the scoring module is configured to score the driving behavior of the current driver using the accumulated sample data using the driving scoring method corresponding to the current accumulation stage.
  • the sample driving data in the accumulated sample data is used to score the driving behavior of the current driver.
  • the driving scoring method corresponding to the second stage is used to score the driving behavior of the current driver using the sample driving data and the sample insurance policy data in the accumulated sample data.
  • the first stage includes a third stage where the accumulated amount of the sample insurance policy data is below the first threshold and the accumulated amount of the sample driving data is below the second threshold,
  • the scoring module is configured to:
  • a rating of the current driver's driving characteristics is obtained as the rating of the current driver's driving behavior.
  • the first stage includes a fourth stage where the accumulation amount of the sample insurance policy data is below the first threshold and the accumulation amount of the sample driving data is above the second threshold,
  • the scoring module is configured to:
  • a cumulative distribution function of the driving characteristics of the current driver is calculated as the driving behavior score of the current driver.
  • the second stage includes the first accumulation of the sample insurance policy data above the first threshold, and the accumulation of the sample insurance policy data is less than the accumulation of the sample driving data Five stages,
  • the scoring module is configured to:
  • the risk level cluster corresponding to the current driver is determined, and the determined driving score of the risk level cluster is used as the current driver's driving score.
  • the second stage includes the accumulation of the sample insurance policy data above the first threshold, and the accumulation of the sample insurance policy data is equal to the third of the accumulation of the sample driving data Six stages,
  • the scoring module is configured to:
  • the driving behavior scoring model is used to score the driving behavior of the current driver.
  • the obtaining driving scores of a plurality of sample drivers corresponding to the policy data according to the sample insurance policy data includes:
  • a driving score having a mapping relationship with the insurance claim value is determined as the driving score of the sample driver.
  • an embodiment of the present disclosure provides a driving behavior scoring device, including:
  • One or more processors are One or more processors;
  • Storage device for storing one or more programs
  • the one or more processors implement the driving behavior scoring method described in the first aspect.
  • an embodiment of the present disclosure provides a computer-readable storage medium on which a computer program is stored, which when executed by a processor implements the driving behavior scoring method described in the first aspect.
  • the driving behavior scoring method, device, and computer-readable storage medium provided by the embodiments of the present disclosure determine the current accumulation stage of the sample data by accumulating sample data and according to the accumulated amount of the sample, and use the driving scoring method corresponding to the current accumulation stage to evaluate the current The driving behavior of the driver is scored. As a result, the driving behavior risk can be evaluated at different stages of data accumulation.
  • FIG. 1 shows an exemplary flowchart of a driving behavior scoring method according to an embodiment of the present disclosure
  • FIG. 2 shows an exemplary flowchart of a driving behavior scoring method according to another embodiment of the present disclosure
  • FIG. 3 shows an exemplary flowchart of a driving behavior scoring method according to another embodiment of the present disclosure
  • FIG. 4 shows an exemplary flowchart of a driving behavior scoring method according to another embodiment of the present disclosure
  • FIG. 5 shows an exemplary flowchart of a driving behavior scoring method according to another embodiment of the present disclosure
  • FIG. 6 shows a block diagram of an exemplary configuration of a driving behavior scoring device according to another embodiment of the present disclosure
  • FIG. 7 shows an exemplary configuration of a computing device that can implement an embodiment of the present disclosure.
  • one way known to the inventors of the present disclosure is to adopt the method of supervised learning in machine learning to obtain the relationship between driving behavior and driving risk.
  • this type of supervised learning method needs to accumulate a large amount of driving data and insurance claim data, and the insurance claim cycle needs to be consistent with the driving data collection cycle. If there is no insurance claim data, or the data does not reach a certain amount, the algorithm will be difficult to start. In actual operation, at the beginning of the operation of the system, it is often more convenient to collect driving data, but insurance and claims usually do not happen immediately, so it is not possible to obtain insurance claims data at the same time as collecting driving data.
  • Embodiments of the present disclosure provide a driving behavior scoring method that can accurately predict driving behavior scores for target drivers by using different driving behavior scoring methods in different stages of data accumulation.
  • the execution subject of the driving behavior scoring method may be a server.
  • the server can communicate with a mobile terminal equipped with a mobile SDK (Software Development Kit) software module and a vehicle depth camera module through a network.
  • the mobile SDK module may be, for example, a software development kit provided by a mobile terminal, which is not limited in this disclosure.
  • the server can also interface with the insurance policy system through a preset interface to obtain driver's insurance policy data.
  • the preset interface involved in the present disclosure refers to some predefined functions in the software system, which are used to interact with other systems, data transmission, etc. The present disclosure does not limit the preset interface.
  • the server may be a single server or a server group composed of multiple servers, and in the server group, multiple servers may communicate and connect.
  • the mobile terminal equipped with the mobile SDK module may be a driver's mobile terminal, and the driver's driving behavior data collected by the mobile SDK module may be uploaded to the server.
  • the vehicle-mounted depth camera module can be installed on the driver's vehicle and can upload the collected driving video data to the server.
  • FIG. 1 shows an exemplary flowchart of a driving behavior scoring method according to an embodiment of the present disclosure.
  • step S1 sample data is accumulated, and the sample data includes sample driving data and sample insurance policy data of a sample driver.
  • the sample driving data of the sample driver may include at least one of driving behavior data, driving environment data, and driving video data.
  • the driving behavior data of the sample driver may be obtained through the mobile SDK module.
  • the driving behavior data can be various basic information when the vehicle is driving, and can include driving time information, mileage information, speed information, steering information, latitude and longitude information, altitude information, mobile phone call status information, rapid acceleration information, rapid deceleration information, and emergency Turn information, etc.
  • the mobile SDK module in the sample driver's mobile terminal can collect GPS, accelerometer and gyro sensor data during the sample driver's driving process to obtain the sample driver's driving behavior data, and The driving behavior data of the sample driver is uploaded to the server, and the server binds and stores the driving behavior data of the sample driver and the identity of the sample driver.
  • the driving environment data of the sample driver may be obtained through a preset interface.
  • the driving environment data may be various environmental information related to driving behavior in the environment in which the driving is located, and may include road speed limits, driving areas, road types, terrain conditions, current weather conditions for driving, and the like.
  • the driving environment data is uploaded to the server, and the server binds and stores the driving environment data of the sample driver and the identity of the sample driver.
  • driving video data may be obtained through a car camera.
  • the driving video data is the video during driving collected by the camera, and may include ranging information, lane information, road condition information, driving events, and the like.
  • the driving video data is uploaded to the server, and the server binds and stores the driving video data of the sample driver and the identity of the sample driver.
  • the sample insurance policy data of the sample driver may be obtained through a preset interface.
  • the insurance policy data can be various information related to the driver’s insurance policy, which can include the driver’s basic information, policy purchase information and number of policy claims, policy claim amount, etc., where the policy purchase information includes the type of insurance and the amount of insurance coverage information.
  • the server may interface with the insurance policy system through a preset interface, and obtain insurance policy data corresponding to the sample driver's identification from the business policy system according to the sample driver's identification.
  • driver identification can be the driver's mobile phone number, user name, ID card or other information that can uniquely identify the driver's identity.
  • step S2 the current accumulation stage of the sample data is determined according to the accumulation amount of the sample driving data and the sample insurance policy data.
  • the accumulation amount of the sample insurance policy data when the accumulation amount of the sample insurance policy data is lower than the first threshold, it is determined that the current accumulation stage of the sample data is the first stage. In addition, when the accumulation amount of the sample insurance policy data is above the first threshold, it is determined that the current accumulation stage of the sample data is the second stage.
  • the driving behavior of the current driver can be scored using only sample driving data. In this way, when there is not enough insurance policy data at the beginning of the sample data collection, the driver's driving behavior can also be scored.
  • the accumulation amount of the sample insurance policy data is sufficient and can be used to score the driving behavior. Therefore, the sample driving data and the sample insurance policy data can be used to score the current driver's driving behavior. In this way, a more optimized driving scoring method can be used to score the driving behavior of the driver.
  • the first stage may include a third stage where the accumulated amount of sample insurance policy data is below the first threshold and the accumulated amount of sample driving data is below the second threshold.
  • the third stage may be a cold start stage where the sample data includes only a small amount of sample driving data and does not include sample insurance policy data.
  • the so-called cold start phase refers to the phase in which the system still starts to operate when specific data has not been obtained in the initial stage of system operation.
  • the cold start stage refers to a stage where a driver's driving behavior is scored based only on sample driving data without any sample insurance policy data.
  • the first stage may include a fourth stage where the accumulated amount of sample insurance policy data is below the first threshold and the accumulated amount of sample driving data is above the second threshold.
  • the fourth stage may be an initial stage in which the sample data includes sample driving data of multiple sample drivers and a very small amount of sample insurance policy data.
  • the second stage may include a fifth stage in which the accumulated amount of sample insurance policy data is above the first threshold and the accumulated amount of sample insurance policy data is less than the accumulated amount of sample driving data.
  • the fifth stage may be a mid-term stage where the sample data includes sample driving data of multiple sample drivers and sample insurance policy data of some sample drivers.
  • the second stage may include a sixth stage in which the accumulated amount of sample insurance policy data is above the first threshold, and the accumulated amount of sample insurance policy data is equal to the accumulated amount of sample driving data.
  • the sixth stage may be a later stage in which the sample data includes sample driving data of multiple sample drivers and corresponding sample insurance policy data.
  • first threshold and the second threshold are predetermined values according to actual conditions, and the disclosure does not limit the way of selecting the threshold.
  • the accumulation amount of the sample insurance policy data is smaller than the accumulation amount of the sample driving data means that the sample data includes sample driving data of a plurality of sample drivers, and only includes a part of the plurality of sample drivers Driver's sample insurance policy data.
  • the accumulation amount of the sample insurance policy data equal to the accumulation amount of the sample driving data means that the sample data includes sample driving data of a plurality of sample drivers, and includes sample insurance policy data of the plurality of sample drivers.
  • the cold start stage, the initial stage, the middle stage, and the late stage are used to represent the third stage, the fourth stage, the fifth stage, and the sixth stage, respectively.
  • step S3 using the driving scoring method corresponding to the current accumulation stage, the accumulated sample data is used to score the driving behavior of the current driver.
  • different driving scoring methods are used in different stages to score driving behaviors.
  • the driving behavior scoring method provided by an embodiment of the present disclosure accumulates sample data and determines the current accumulation stage of the sample data, and uses the driving scoring method corresponding to the current accumulation stage to score the current driver's driving behavior.
  • driving behavior scores can be provided at the beginning of driving data collection, and with the accumulation of driving data and the addition of claim data, parameters and models are gradually optimized, so that driving behavior risks can be carried out at different stages of data accumulation Assessment.
  • a driving scoring method based on rules and preset weights may be used Rate the driving behavior of the current driver.
  • using the driving scoring method corresponding to the current accumulation stage to score the driving behavior of the current driver in the foregoing step S3 may include the steps of:
  • the current driver's driving data may include at least one of driving behavior data, driving environment data, and driving video data.
  • the driving behavior data can be obtained through the mobile SDK module, and the driving mileage, driving duration, maximum driving speed, number of sharp accelerations, sharp decelerations, and sharp turns including each journey or unit time period can be extracted from the driving behavior data At least one driving behavior characteristic of the number of times, the number of sharp lane changes, whether to drive fatigued, and whether to drive during a dangerous period.
  • the driving video data can be obtained through the on-board camera, and at least one driving video feature including distance to vehicle speed ratio, whether to follow the lane, and whether to decelerate when the front event occurs is extracted from the driving video data.
  • the driving characteristics are divided into several dimensions, and each dimension includes several driving characteristics.
  • the extracted multiple driving features can be divided into several dimensions, such as speed, mileage, time, environment, driving behavior, etc.
  • Each dimension contains several features.
  • the "speed” dimension can include several features related to speed: the number of speeding, the length of overspeed, the maximum driving speed, etc.
  • the "driving behavior” dimension can include several features related to driving operations: the number of rapid accelerations, The number of sharp decelerations, the number of sharp turns, the number of sharp lane changes, the distance speed ratio, whether to follow the lane, and whether to slow down when an event in front occurs.
  • scoring standards can be set for various driving characteristics, and the scoring standards corresponding to various driving characteristics can be set differently according to specific circumstances.
  • the scoring criteria for the driving characteristic of the number of speeding can be:
  • the scores of each feature in each dimension can be weighted and summed to obtain the score of each dimension, where the sum of the feature weights in each dimension is 1.
  • the driving behavior score of each driver can be obtained by weighting and summing the scores of each dimension according to the dimension weights set for each dimension, where the sum of the weights of each dimension is 1.
  • driving behavior score ⁇ dimension score * dimension weight.
  • the driving characteristics of the current driver may not be dimensionally divided, and after obtaining the driving characteristics score, the driving behavior score of the current driver may be directly used .
  • the driving scoring method based on the feature distribution function can be used to score the driving behavior of the current driver.
  • using the driving scoring method corresponding to the current accumulation stage to score the driving behavior of the current driver in the foregoing step S3 may include the steps of:
  • step S311 which will not be repeated here.
  • S322. Calculate the distribution parameters of various sample driving characteristics according to one or more sample driving characteristics of each sample driver.
  • the sample driving characteristics of all sample drivers can be aggregated and the aggregated data can be cleaned, including necessary outlier detection and data normalization, based on the cleaned
  • the data fitting obtains the parameters of the distribution function that the driving characteristics of the sample obey, and so on to obtain the distribution parameters of the driving characteristics of other samples.
  • step S311 which will not be repeated here.
  • the cumulative distribution function value of the driving characteristic is obtained according to the exponentially distributed parameter ⁇ acc Take this as the score corresponding to the driving feature.
  • S325. Perform weighted calculation on the scores of various driving characteristics to obtain the current driver's driving behavior score.
  • the current stage is the initial stage
  • a certain amount of driving data of the sample driver is accumulated, and the insurance policy data of the sample driver is still little or little, and the overall correlation with the driving data of the driver is relatively low
  • the current stage is an intermediate stage in the accumulation process of sample data
  • a certain amount of driving data of a sample driver and insurance policy data of some sample drivers are accumulated, and clustering-based
  • the analyzed driving scoring method scores the current driver's driving behavior.
  • the driving scoring method corresponding to the current accumulation stage is used to score the driving behavior of the current driver, which may specifically include the steps of:
  • the insurance claim value from the insurance policy data, and determine the driving score that has a mapping relationship with the insurance claim value according to the preset mapping relationship table as the driving score of the sample driver; where, in the preset mapping relationship table, the claim rate The higher the corresponding driving behavior score, the lower the score.
  • the insurance claim value can be extracted from the insurance data of the sample driver’s insurance policy data, and the driving risk score of the sample driver can be obtained from the preset mapping relationship table according to the insurance rate or claim rate, Use as a rating label for the sample driver.
  • the mapping relationship table the higher the claim rate, the lower the driving behavior score.
  • the score range may be 0-100.
  • the K-means algorithm can be used to perform cluster analysis on the driving data of some sample drivers.
  • Each category is used as a risk level cluster to obtain a cluster center, and it is considered that drivers in the same category have the same or similar risk levels.
  • the average of the driving scores of the sample drivers in the same category is taken as the driving score of the category (ie, the risk level cluster).
  • S333 Obtain the current driver's driving data, and calculate the similarity distance between the current driver's driving data and multiple cluster centers.
  • the embodiments of the present disclosure do not limit the specific calculation process.
  • the risk level cluster corresponding to the minimum value in the calculation result of the similarity distance is taken as the risk level cluster corresponding to the current driver.
  • the scoring method scores the current driver's driving behavior, which can solve the problem in the prior art that due to insufficient accumulation of sample data in the initial stage of system operation, it is difficult to start the supervision algorithm, which leads to the inability to score the driving behavior.
  • the sample driver’s driving data and claim data are relatively complete, and the driver’s driving data and claim data are relatively well correlated.
  • a driving score based on the trained scoring model can be used Methods Score the driving behavior of the current driver. Specifically, as shown in FIG. 5, the aforementioned step S3 uses the driving scoring method corresponding to the current accumulation stage to score the driving behavior of the current driver, which may specifically include the steps of:
  • step S331 the process of obtaining driving scores of a plurality of sample drivers corresponding to the sample insurance policy data according to the sample insurance policy data of the sample driver is the same as step S331, and details are not described here.
  • step S311 the process of extracting the sample driving characteristics of the plurality of sample drivers from the sample driving data is the same as step S311, and will not be repeated here.
  • the driving characteristics of each sample driver and the driving score of each sample driver are established through traditional machine learning or deep learning methods, and the driving behavior scoring model is stored offline for online driving behavior scoring Call.
  • linear regression random forest, decision tree, xgboost and other methods can be used to establish a driving behavior scoring model, which is not limited in this disclosure.
  • the current driver's driving characteristics are extracted from the current driver's driving data; the current driver's driving characteristics are input into the driving behavior scoring model, and the current driver's driving behavior score is obtained and output.
  • machine learning or Deep learning training obtains a driving scoring model, and then scores the driving behavior of the current driver.
  • FIG. 6 shows a block diagram of an exemplary configuration of a driving behavior scoring device according to another embodiment of the present disclosure.
  • the driving behavior scoring device provided by the embodiment of the present disclosure may be used to execute the driving behavior scoring method in the above embodiment.
  • the device 600 may include a processing circuit 60.
  • the processing circuit 60 of the device 600 provides various functions of the device 600.
  • the processing circuit 60 of the device 600 may be configured to perform the driving behavior scoring method described above with reference to FIG. 1.
  • the processing circuit 60 may refer to various implementations of digital circuitry, analog circuitry, or mixed-signal (combination of analog and digital) circuitry that performs functions in a computing system.
  • the processing circuit may include, for example, a circuit such as an integrated circuit (IC), an application specific integrated circuit (ASIC), a part or circuit of a separate processor core, an entire processor core, a separate processor, such as a field programmable gate array (FPGA) Programmable hardware devices, and/or systems that include multiple processors.
  • IC integrated circuit
  • ASIC application specific integrated circuit
  • FPGA field programmable gate array
  • the processing circuit 60 may include an accumulation module 61, a determination module 62, and a scoring module 63.
  • the accumulation module 61 is configured to accumulate sample data including the sample driving data and sample insurance policy data of the sample driver; the determination module 62 is configured to determine based on the accumulated amount of the sample driving data and sample insurance policy data The current accumulation stage of the sample data; the scoring module 63 is configured to score the current driver's driving behavior using the accumulated sample data using the driving scoring method corresponding to the current accumulation stage.
  • the above-mentioned modules 61 to 63 may be configured to perform steps S1 to S3 in the aforementioned driving behavior scoring method shown in FIG. 1, respectively.
  • the device 600 may further include a memory (not shown).
  • the memory of the device 600 may store the information generated by the processing circuit 60 and the programs and data used for the operation of the device 600.
  • the memory may be volatile memory and/or non-volatile memory.
  • the memory may include, but is not limited to, random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), read only memory (ROM), and flash memory.
  • the apparatus 600 may be implemented at the chip level, or may also be implemented at the device level by including other external components.
  • modules are only logical modules divided according to the specific functions they implement, and are not intended to limit specific implementations. In actual implementation, the above modules may be implemented as independent physical entities, or may be implemented by a single entity (for example, a processor (CPU or DSP, etc.), an integrated circuit, etc.).
  • the driving behavior scoring device provided by the embodiment of the present disclosure and the driving behavior scoring method provided by the embodiments of the present disclosure belong to the same inventive concept, and can execute the driving behavior scoring method provided by any embodiment of the present disclosure, and have the corresponding Functional modules and beneficial effects.
  • the driving behavior scoring method provided in the embodiments of the present disclosure and details are not repeated here.
  • FIG. 7 shows an exemplary configuration of a computing device 700 that can implement embodiments according to the present disclosure.
  • the computing device 700 is an example of a hardware device to which the above-mentioned aspects of the present disclosure can be applied.
  • the computing device 700 may be any machine configured to perform processing and/or calculations.
  • the computing device 700 may be, but not limited to, a workstation, server, desktop computer, laptop computer, tablet computer, personal data assistant (PDA), smart phone, in-vehicle computer, or combination thereof.
  • PDA personal data assistant
  • the computing device 700 may include one or more elements that may be connected to or communicate with the bus 702 via one or more interfaces.
  • the bus 702 may include, but is not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, Enhanced ISA (EISA) bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus, etc.
  • the computing device 700 may include, for example, one or more processors 704, one or more input devices 706, and one or more output devices 708.
  • the one or more processors 704 may be any kind of processors, and may include, but are not limited to, one or more general-purpose processors or special-purpose processors (such as special-purpose processing chips).
  • the processor 704 may correspond to, for example, the processing circuit 60 in FIG. 6, and is configured to implement the functions of each module of the verification device of the present disclosure and the certificate holder.
  • the input device 706 may be any type of input device capable of inputting information to a computing device, and may include, but is not limited to, a mouse, keyboard, touch screen, microphone, and/or remote controller.
  • the output device 708 may be any type of device capable of presenting information, and may include, but is not limited to, a display, a speaker, a video/audio output terminal, a vibrator, and/or a printer.
  • the computing device 700 may also include or be connected to a non-transitory storage device 714, which may be any non-transitory storage device that can implement data storage, and may include but not limited to disk drives, optical Storage devices, solid-state memory, floppy disks, flexible disks, hard disks, magnetic tape, or any other magnetic media, compact disks, or any other optical media, cache memory, and/or any other memory chips or modules, and/or computers can read data , Instructions and/or any other medium of code.
  • the computing device 700 may also include random access memory (RAM) 710 and read-only memory (ROM) 712.
  • the ROM 712 may store programs, utilities, or processes to be executed in a non-volatile manner.
  • RAM 710 may provide volatile data storage and store instructions related to the operation of computing device 700.
  • the computing device 700 may also include a network/bus interface 716 coupled to the data link 718.
  • the network/bus interface 716 may be any kind of device or system capable of enabling communication with external devices and/or networks, and may include, but is not limited to, a modem, a network card, an infrared communication device, a wireless communication device, and/or a chipset (such as Bluetooth (TM) device, an 802.11 device, WiFi equipment, WiMax, cellular communication facilities, etc.).
  • TM Bluetooth
  • another embodiment of the present disclosure also provides a driving behavior scoring device, including:
  • One or more processors are One or more processors;
  • Storage device for storing one or more programs
  • the one or more processors implement the driving behavior scoring method described in the foregoing embodiment.
  • another embodiment of the present disclosure provides a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the driving behavior scoring method described in the above embodiment is implemented.
  • the embodiments in the embodiments of the present disclosure may be provided as methods, systems, or computer program products. Therefore, the embodiments of the present disclosure may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Moreover, the embodiments of the present disclosure may take the form of computer program products implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program code .
  • computer usable storage media including but not limited to disk storage, CD-ROM, optical storage, etc.
  • These computer program instructions can be provided to the processor of a general-purpose computer, special-purpose computer, embedded processing machine, or other programmable data processing device to produce a machine that enables the generation of instructions executed by the processor of the computer or other programmable data processing device
  • These computer program instructions may also be stored in a computer-readable memory that can guide a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction device, the instructions
  • the device implements the functions specified in one block or multiple blocks of the flowchart one flow or multiple flows and/or block diagrams.
  • These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operating steps are performed on the computer or other programmable device to produce computer-implemented processing, which is executed on the computer or other programmable device
  • the instructions provide steps for implementing the functions specified in one block or multiple blocks of the flowchart one flow or multiple flows and/or block diagrams.

Landscapes

  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Engineering & Computer Science (AREA)
  • Strategic Management (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • Educational Administration (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Game Theory and Decision Science (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Technology Law (AREA)
  • Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
  • Control Of Driving Devices And Active Controlling Of Vehicle (AREA)

Abstract

一种驾驶行为评分方法、装置及计算机可读存储介质。驾驶行为评分方法包括：积累样本数据，所述样本数据包括样本驾驶员的样本驾驶数据及样本保险保单数据（S1）；根据所述样本驾驶数据及样本保险保单数据的积累量，确定所述样本数据的当前积累阶段（S2）；使用所述当前积累阶段对应的驾驶评分方法，利用所积累的样本数据对当前驾驶员的驾驶行为进行评分（S3）。

Description

一种驾驶行为评分方法、装置及计算机可读存储介质 技术领域
本公开涉及车辆驾驶行为分析技术领域，尤其涉及一种驾驶行为评分方法、装置及计算机可读存储介质。
背景技术
近年来，机动车辆的数量越来越多，同时随着车联网飞速发展，与车联网技术结合的基于驾驶员驾驶行为进行定价的车险产品也越来越多。
UBI(Usage Based Insurance，基于使用量的保险)车险是指基于驾驶行为以及车辆使用数据定价的车辆保险，是在车险费率自由化的背景下出现的新型车险。保险公司通过收集驾驶行为数据以及车辆运行数据，经过数据分析处理，评估驾驶人员的驾驶行为风险等级。对于驾驶行为好的车主，保险公司可以给予较多优惠，对于驾驶行为不好的车主，保险公司可以相应地提高保费。由此，有助于车主培养良好的驾驶行为，减少事故发生，减少出险与理赔，从而改善交通状况，降低社会成本。
因此，对于保险公司，在为车险产品进行定价前首先需要基于驾驶人员的驾驶数据作出驾驶行为风险的评估，目前通常是采用机器学习中监督学习的方法，获得驾驶行为与驾驶风险之间的关系，训练样本需要同时包含驾驶行为和车辆使用数据以及出险理赔数据。采集驾驶行为以及车辆使用数据，通常通过车载OBD(On-Board Diagnostic，车载诊断)设备或者智能手机等设备，由设备内的传感器获得车辆行驶数据，如果设备附带摄像头，还可以由摄像头获取行驶视频从而得到体现驾驶习惯的进一步数据。作为风险程度的准确衡量，需要知道被观测的驾驶人员在驾驶行为观测的同一周期内是否发生交通事故，以及向保险公司索赔的情况，以出险率或者赔付率作为驾驶风险的表征。在获得以上两者以后，驾驶数据作为监督算法中的特征，出险率或赔付率作为监督算法中的监督信号，监督算法基于训练数据得到特征与监督信号的关系，也就是驾驶数据与出险率、理赔率的关系。对于需要判断驾驶行为风险的目标样本，也就是只有驾驶数据的驾驶人员，由监督算法得到的推断关系获得其驾驶行为风险。
发明内容
在下文中给出了关于本公开的简要概述，以便提供关于本公开的一些方面的基本理解。但是，应当理解，这个概述并不是关于本公开的穷举性概述。它并不是意图用来确定本公 开的关键性部分或重要部分，也不是意图用来限定本公开的范围。其目的仅仅是以简化的形式给出关于本公开的某些概念，以此作为稍后给出的更详细描述的前序。
本公开实施例的目的是提供一种驾驶行为评分方法、装置及计算机可读存储介质，在数据积累的不同阶段中，通过使用不同的驾驶评分方法对目标驾驶员准确地进行驾驶行为评分预测。本公开实施例的具体技术方案如下：
第一方面，本公开实施例提供了一种驾驶行为评分方法，所述方法包括：
积累样本数据，所述样本数据包括样本驾驶员的样本驾驶数据及样本保险保单数据；
根据所述样本驾驶数据及样本保险保单数据的积累量，确定所述样本数据的当前积累阶段；
使用所述当前积累阶段对应的驾驶评分方法，利用所积累的样本数据对当前驾驶员的驾驶行为进行评分。
在一些实施例中，在所述样本保险保单数据的积累量低于第一阈值时，确定所述样本数据的当前积累阶段为第一阶段，
使用所述第一阶段对应的驾驶评分方法，利用所积累的样本数据中的样本驾驶数据对所述当前驾驶员的驾驶行为进行评分。
在一些实施例中，在所述样本保险保单数据的积累量在第一阈值以上时，确定所述样本数据的当前积累阶段为第二阶段，
使用所述第二阶段对应的驾驶评分方法，利用所积累的样本数据中的样本驾驶数据及样本保险保单数据对所述当前驾驶员的驾驶行为进行评分。
在一些实施例中，所述第一阶段包括所述样本保险保单数据的积累量低于所述第一阈值并且所述样本驾驶数据的积累量低于第二阈值的第三阶段，
在所述当前积累阶段为所述第三阶段时，所述对当前驾驶员的驾驶行为进行评分包括：
获取所述当前驾驶员的驾驶数据，并提取所述当前驾驶员的驾驶特征；
根据预设评分标准，获取所述当前驾驶员的驾驶特征的评分，作为所述当前驾驶员的驾驶行为评分。
在一些实施例中，所述第一阶段包括所述样本保险保单数据的积累量低于所述第一阈值并且所述样本驾驶数据的积累量在第二阈值以上的第四阶段，
在所述当前积累阶段为所述第四阶段时，所述对当前驾驶员的驾驶行为进行评分包括：
从所述样本驾驶员的样本驾驶数据中提取所述样本驾驶员的样本驾驶特征；
根据所述样本驾驶员的样本驾驶特征，计算所述样本驾驶特征的分布参数；
获取所述当前驾驶员的驾驶数据，并提取所述当前驾驶员的驾驶特征；
根据所述样本驾驶特征的分布参数，计算所述当前驾驶员的驾驶特征的累积分布函数，作为所述当前驾驶员的驾驶行为评分。
在一些实施例中，所述第二阶段包括所述样本保险保单数据的积累量在所述第一阈值以上、并且所述样本保险保单数据的积累量小于所述样本驾驶数据的积累量的第五阶段，
在所述当前积累阶段为所述第五阶段时，所述对当前驾驶员的驾驶行为进行评分包括：
根据所述样本保险保单数据，获取与该样本保险保单数据对应的多个样本驾驶员的驾驶评分；
对所述多个样本驾驶员的驾驶数据进行聚类分析，获取多个聚类中心，每一类作为一个风险等级簇，并根据所述多个样本驾驶员的驾驶评分，计算各个风险等级簇的驾驶评分；
获取所述当前驾驶员的驾驶数据，并计算所述当前驾驶员的驾驶数据与所述多个聚类中心的相似度距离；
根据所述相似度距离的计算结果，确定所述当前驾驶员所对应的风险等级簇，并将确定出的所述风险等级簇的驾驶评分作为所述当前驾驶员的驾驶评分。
在一些实施例中，所述第二阶段包括所述样本保险保单数据的积累量在所述第一阈值以上、并且所述样本保险保单数据的积累量等于所述样本驾驶数据的积累量的第六阶段，
在所述当前积累阶段为所述第六阶段时，所述对当前驾驶员的驾驶行为进行评分包括：
根据所述样本保险保单数据，获取与该样本保险保单数据对应的多个样本驾驶员的驾驶评分；
从所述样本驾驶数据中提取所述多个样本驾驶员的样本驾驶特征；
根据所述多个样本驾驶员的样本驾驶特征及对应的驾驶评分，建立驾驶行为评分模型；
使用所述驾驶行为评分模型对所述当前驾驶员的驾驶行为进行评分。
在一些实施例中，所述根据所述样本保险保单数据，获取与该保单数据对应的多个样本驾驶员的驾驶评分包括：
针对各所述多个样本驾驶员的样本保险保单数据，执行以下操作：
从所述样本保险保单数据中获取出险理赔数值；
根据预设映射关系表，确定与所述出险理赔数值具有映射关系的驾驶评分，以作为所述样本驾驶员的驾驶评分。
第二方面，本公开实施例提供一种驾驶行为评分装置，所述装置包括：
积累模块，被配置为积累样本数据，所述样本数据包括样本驾驶员的样本驾驶数据及样本保险保单数据；
确定模块，被配置为根据所述样本驾驶数据及样本保险保单数据的积累量，确定所述 样本数据的当前积累阶段；
评分模块，被配置为使用所述当前积累阶段对应的驾驶评分方法，利用所积累的样本数据对当前驾驶员的驾驶行为进行评分。
在一些实施例中，在所述样本保险保单数据的积累量低于第一阈值时，确定所述样本数据的当前积累阶段为第一阶段，
使用所述第一阶段对应的驾驶评分方法，利用所积累的样本数据中的样本驾驶数据对所述当前驾驶员的驾驶行为进行评分。
在一些实施例中，在所述样本保险保单数据的积累量在第一阈值以上时，确定所述样本数据的当前积累阶段为第二阶段，
使用所述第二阶段对应的驾驶评分方法，利用所积累的样本数据中的样本驾驶数据及样本保险保单数据对所述当前驾驶员的驾驶行为进行评分。
在一些实施例中，所述第一阶段包括所述样本保险保单数据的积累量低于所述第一阈值并且所述样本驾驶数据的积累量低于第二阈值的第三阶段，
在所述当前积累阶段为所述第三阶段时，所述评分模块被配置为：
获取所述当前驾驶员的驾驶数据，并提取所述当前驾驶员的驾驶特征；
根据预设评分标准，获取所述当前驾驶员的驾驶特征的评分，作为所述当前驾驶员的驾驶行为评分。
在一些实施例中，所述第一阶段包括所述样本保险保单数据的积累量低于所述第一阈值并且所述样本驾驶数据的积累量在第二阈值以上的第四阶段，
在所述当前积累阶段为所述第四阶段时，所述评分模块被配置为：
从所述样本驾驶员的样本驾驶数据中提取所述样本驾驶员的样本驾驶特征；
根据所述样本驾驶员的样本驾驶特征，计算所述样本驾驶特征的分布参数；
获取所述当前驾驶员的驾驶数据，并提取所述当前驾驶员的驾驶特征；
根据所述样本驾驶特征的分布参数，计算所述当前驾驶员的驾驶特征的累积分布函数，作为所述当前驾驶员的驾驶行为评分。
在一些实施例中，所述第二阶段包括所述样本保险保单数据的积累量在所述第一阈值以上、并且所述样本保险保单数据的积累量小于所述样本驾驶数据的积累量的第五阶段，
在所述当前积累阶段为所述第五阶段时，所述评分模块被配置为：
根据所述样本保险保单数据，获取与该样本保险保单数据对应的多个样本驾驶员的驾驶评分；
对所述多个样本驾驶员的驾驶数据进行聚类分析，获取多个聚类中心，每一类作为一 个风险等级簇，并根据所述多个样本驾驶员的驾驶评分，计算各个风险等级簇的驾驶评分；
获取所述当前驾驶员的驾驶数据，并计算所述当前驾驶员的驾驶数据与所述多个聚类中心的相似度距离；
根据所述相似度距离的计算结果，确定所述当前驾驶员所对应的风险等级簇，并将确定出的所述风险等级簇的驾驶评分作为所述当前驾驶员的驾驶评分。
在一些实施例中，所述第二阶段包括所述样本保险保单数据的积累量在所述第一阈值以上、并且所述样本保险保单数据的积累量等于所述样本驾驶数据的积累量的第六阶段，
在所述当前积累阶段为所述第六阶段时，所述评分模块被配置为：
根据所述样本保险保单数据，获取与该样本保险保单数据对应的多个样本驾驶员的驾驶评分；
从所述样本驾驶数据中提取所述多个样本驾驶员的样本驾驶特征；
根据所述多个样本驾驶员的样本驾驶特征及对应的驾驶评分，建立驾驶行为评分模型；
使用所述驾驶行为评分模型对所述当前驾驶员的驾驶行为进行评分。
在一些实施例中，所述根据所述样本保险保单数据，获取与该保单数据对应的多个样本驾驶员的驾驶评分包括：
针对各所述多个样本驾驶员的样本保险保单数据，执行以下操作：
从所述样本保险保单数据中获取出险理赔数值；
根据预设映射关系表，确定与所述出险理赔数值具有映射关系的驾驶评分，以作为所述样本驾驶员的驾驶评分。
第三方面，本公开实施例提供一种驾驶行为评分装置，包括：
一个或多个处理器；
存储装置，用于存储一个或多个程序；
当所述一个或多个程序被所述一个或多个处理器执行，使得所述一个或多个处理器实现如第一方面所述的驾驶行为评分方法。
第四方面，本公开实施例提供一种计算机可读存储介质，其上存储有计算机程序，所述程序被处理器执行时实现如第一方面所述的驾驶行为评分方法。本公开实施例提供的驾驶行为评分方法、装置及计算机可读存储介质，通过积累样本数据，并根据样本的积累量确定样本数据的当前积累阶段，以及使用当前积累阶段对应的驾驶评分方法对当前驾驶员的驾驶行为进行评分。由此，在数据积累的不同阶段均可进行驾驶行为风险的评估。
附图说明
为了更清楚地说明本公开实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本公开的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1示出了根据本公开一实施例的驾驶行为评分方法的示例性流程图；
图2示出了根据本公开另一实施例的驾驶行为评分方法的示例性流程图；
图3示出了根据本公开另一实施例的驾驶行为评分方法的示例性流程图；
图4示出了根据本公开另一实施例的驾驶行为评分方法的示例性流程图；
图5示出了根据本公开另一实施例的驾驶行为评分方法的示例性流程图；
图6示出了根据本公开另一实施例的驾驶行为评分装置的示例性配置框图；
图7示出了可以实现根据本公开的实施例的计算设备的示例性配置。
具体实施方式
为使本公开的目的、技术方案和优点更加清楚，下面将结合本公开实施例中的附图，对本公开实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本公开一部分实施例，而不是全部的实施例。基于本公开中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本公开保护的范围。
在基于驾驶人员的驾驶数据作出驾驶行为风险的评估时，本公开的发明人知晓的一种方式是采用机器学习中监督学习的方法，获得驾驶行为与驾驶风险之间的关系。然而，这一类监督学习的方法需要积累大量的驾驶数据与出险理赔数据，而且出险理赔周期需要与驾驶数据采集周期一致。如果没有出险理赔数据，或者数据达不到一定数量，算法就难以启动。而在实际操作中，在系统运行之初，往往是驾驶数据比较方便采集，但出险与理赔通常不会马上发生，所以并不能在采集驾驶数据的同时就获得出险理赔数据。同时随着驾驶行程与数据的积累，即使发生出险，由于理赔流程包含查勘、定损、维修、理赔等多个步骤，需要一定时间，出险以及理赔数据的获取可能还存在一段时间的延迟，进而使得监督算法也难以启动。
本公开实施例提供了一种驾驶行为评分方法，该方法能够在数据积累的不同阶段中通过使用不同的驾驶行为评分方法对目标驾驶员准确地进行驾驶行为评分预测。
该驾驶行为评分方法的执行主体可以是服务器。服务器可以通过网络与搭载有移动SDK(Software Development Kit，软件开发工具包)模块的移动终端及车载深度摄像头模块进行通信连接。移动SDK模块例如可以是移动终端提供的软件开发工具包，本公开对 此不作限定。
服务器还可以通过预设接口与保险保单系统进行对接，以获取驾驶员的保险保单数据。本公开所涉及的预设接口是指软件系统中一些预先定义的函数，用于与其它系统进行交互、数据传输等，本公开对预设接口没有限定。
服务器可以是单个服务器，也可以是由多个服务器组成的服务器群，且该服务器群内，多个服务器之间可以进行通信连接。搭载有移动SDK模块的移动终端可以为驾驶员的移动终端，其可以通过移动SDK模块采集到驾驶员的驾驶行为数据上传至服务器。车载深度摄像头模块可以安装在驾驶员的车辆上，能够将采集到的驾驶视频数据上传至服务器。
图1示出了根据本公开一实施例的驾驶行为评分方法的示例性流程图。
如图1所示，在步骤S1中，积累样本数据，所述样本数据包括样本驾驶员的样本驾驶数据及样本保险保单数据。
在一些实施例中，样本驾驶员的样本驾驶数据可以包括驾驶行为数据、驾驶环境数据和驾驶视频数据中的至少一种。
在一些实施例中，可以通过移动SDK模块获取样本驾驶员的驾驶行为数据。驾驶行为数据可以是车辆行驶时的各种基本信息，可以包括驾驶时间信息、里程信息、速度信息、转向信息、经纬度信息、海拔信息、手机通话状态信息、急加速信息、急减速信息、以及急转弯信息等。
在具体实施过程中，可以通过样本驾驶员的移动终端内的移动SDK模块在样本驾驶员的驾驶过程中采集GPS、加速度计和陀螺仪传感器数据，以获取样本驾驶员的驾驶行为数据，并将样本驾驶员的驾驶行为数据上传到服务器，由服务器对样本驾驶员的驾驶行为数据与样本驾驶员的身份标识进行绑定存储。
在一些实施例中，可以通过预设接口获取样本驾驶员的驾驶环境数据。驾驶环境数据可以是驾驶所处的环境中各种与驾驶行为相关的环境信息，可以包括路段限速、行车区域、道路类型、地形条件、驾驶当前天气状况等。当通过预设接口获取样本驾驶员的驾驶环境数据后，将驾驶环境数据上传到服务器，由服务器对样本驾驶员的驾驶环境数据与样本驾驶员的身份标识进行绑定存储。
在一些实施例中，可以通过车载摄像头获取驾驶视频数据。驾驶视频数据是摄像头采集到的驾驶期间的视频，可以包括测距信息、车道信息、路况信息、行车事件等。当车载摄像头采集到驾驶视频数据后，将驾驶视频数据上传到服务器，由服务器对样本驾驶员的驾驶视频数据与样本驾驶员的身份标识进行绑定存储。
在一些实施例中，可以通过预设接口获取样本驾驶员的样本保险保单数据。保险保单 数据可以是与驾驶员的保险保单相关的各种信息，可以包括驾驶员的基本信息、保单购买信息和保单理赔单数、保单理赔额等，其中，保单购买信息包括投保险种、投保额度等信息。
在具体实施过程中，服务器可以通过预设接口与保险保单系统进行对接，根据样本驾驶员的身份标识，从业务保单系统中获取到与样本驾驶员的身份标识对应的保险保单数据。
需要说明的是，上述的驾驶员身份标识可以是驾驶员的手机号、用户名、身份证或其他能够唯一标识驾驶员身份的信息。
在步骤S2中，根据所述样本驾驶数据及样本保险保单数据的积累量，确定所述样本数据的当前积累阶段。
在一些实施例中，在样本保险保单数据的积累量低于第一阈值时，确定样本数据的当前积累阶段为第一阶段。另外，在样本保险保单数据的积累量在第一阈值以上时，确定样本数据的当前积累阶段为第二阶段。
当样本数据的当前积累阶段为第一阶段时，样本保险保单数据的积累量较低，不足以用于对驾驶行为进行评分。因此，可以仅利用样本驾驶数据对当前驾驶员的驾驶行为进行评分。通过这种方式，在样本数据采集之初还没有足够的保险保单数据的情况下，也能够对驾驶员的驾驶行为进行评分。
当样本数据的当前积累阶段为第二阶段时，样本保险保单数据的积累量较充足，可以用于对驾驶行为进行评分。因此，可以同时利用样本驾驶数据以及样本保险保单数据对当前驾驶员的驾驶行为进行评分。通过这种方式，能够采用更优化的驾驶评分方法对驾驶员的驾驶行为进行评分。
在一些实施例中，第一阶段可以包括样本保险保单数据的积累量低于第一阈值并且样本驾驶数据的积累量低于第二阈值的第三阶段。例如，第三阶段可以是样本数据仅包括少量样本驾驶数据而不包括样本保险保单数据的冷启动阶段。所谓冷启动阶段，即在系统运行初期，尚未获得特定数据的情况下仍然启动系统运行的阶段。在本公开中，冷启动阶段是指在没有任何样本保险保单数据的情况下仅基于样本驾驶数据对驾驶员的驾驶行为进行评分的阶段。
在一些实施例中，第一阶段可以包括样本保险保单数据的积累量低于第一阈值并且样本驾驶数据的积累量在第二阈值以上的第四阶段。例如，第四阶段可以是样本数据包括多个样本驾驶员的样本驾驶数据和极少量的样本保险保单数据的初期阶段。
在一些实施例中，第二阶段可以包括样本保险保单数据的积累量在所述第一阈值以上、并且样本保险保单数据的积累量小于样本驾驶数据的积累量的第五阶段。例如，第五阶段 可以是样本数据包括多个样本驾驶员的样本驾驶数据和部分样本驾驶员的样本保险保单数据的中期阶段。
在一些实施例中，第二阶段可以包括样本保险保单数据的积累量在第一阈值以上、并且样本保险保单数据的积累量等于样本驾驶数据的积累量的第六阶段。例如，第六阶段可以是样本数据包括多个样本驾驶员的样本驾驶数据和对应的样本保险保单数据的后期阶段。
应当理解，上述第一阈值和第二阈值是根据实际情况预先确定的值，本公开对于阈值的选取方式没有限定。
另外，在本公开中，样本保险保单数据的积累量小于样本驾驶数据的积累量是指，样本数据中包括多个样本驾驶员的样本驾驶数据，而仅包括该多个样本驾驶员中的一部分驾驶员的样本保险保单数据。另外，样本保险保单数据的积累量等于样本驾驶数据的积累量是指，样本数据中包括多个样本驾驶员的样本驾驶数据，并且包括该多个样本驾驶员的样本保险保单数据。
在以下的描述中，使用冷启动阶段、初期阶段、中期阶段和后期阶段分别代表第三阶段、第四阶段、第五阶段和第六阶段进行说明。
在步骤S3中，使用所述当前积累阶段对应的驾驶评分方法，利用所积累的样本数据对当前驾驶员的驾驶行为进行评分。其中，不同阶段使用不同的驾驶评分方法进行驾驶行为评分。
本公开实施例提供的驾驶行为评分方法，通过积累样本数据，并确定样本数据的当前积累阶段，以及使用当前积累阶段对应的驾驶评分方法对当前驾驶员的驾驶行为进行评分。由此可以在驾驶数据采集之初就能够提供驾驶行为评分，并且随着驾驶数据的积累以及理赔数据的加入，逐渐优化参数与模型，从而实现在数据积累的不同阶段均可进行驾驶行为风险的评估。
在本公开的一个实施例中，在当前阶段为冷启动阶段时，样本驾驶员的驾驶数据刚刚开始积累，并没有样本驾驶员的保险保单数据，可以使用基于规则与预设权重的驾驶评分方法对当前驾驶员进行驾驶行为评分。具体的，如图2所示，前述步骤S3中使用当前积累阶段对应的驾驶评分方法对当前驾驶员的驾驶行为进行评分，可以包括步骤：
S311、获取当前驾驶员的驾驶数据，并提取当前驾驶员的一种或多种驾驶特征。
其中，当前驾驶员的驾驶数据可以包括驾驶行为数据、驾驶环境数据和驾驶视频数据中的至少一种。
具体来说，可以通过移动SDK模块获取驾驶行为数据，并从驾驶行为数据中提取包 括每段行程或者单位时间段的驾驶里程、驾驶时长、最大驾驶速度、急加速次数、急减速次数、急转弯次数、急变道次数、是否疲劳驾驶、是否在危险时段驾驶中的至少一个驾驶行为特征。
可以通过预设接口获取驾驶环境数据，并从驾驶环境数据中提取包括每段行程或者单位时间段的超速次数与时长、是否处于危险环境、是否处于熟悉路段、是否有恶劣天气中的至少一个驾驶环境特征。
可以通过车载摄像头获取驾驶视频数据，从驾驶视频数据中提取包括距离车速比、是否遵循车道、前方事件发生时是否减速中的至少一个驾驶视频特征。
S312、对驾驶特征划分成若干维度，各维度包含若干种驾驶特征。
具体的，可以将提取的多种驾驶特征划分为若干维度，如速度、里程、时间、环境、驾驶行为等，每个维度包含若干特征。例如，“速度”这一维度可以包含速度相关的几个特征：超速次数、超速时长、最高驾驶速度等，“驾驶行为”这一维度可以包含与驾驶操作相关的几个特征：急加速次数、急减速次数、急转弯次数、急变道次数、距离车速比、是否遵循车道、前方事件发生时是否减速等。
S313、根据预设评分标准，获取各种驾驶特征的评分。
具体的，可以对各种驾驶特征分别设定评分标准，各种驾驶特征对应的评分标准可以根据具体情况做不同设定。
示例性地，超速次数这一驾驶特征的评分标准可以为：
超速次数得分＝100–超速次数*10/驾驶里程，取下界为0，表示平均每公里超速1次则扣减10分，平均每公里超速次数>＝10，该驾驶特征得分为0。
S314、将各个维度内的各种驾驶特征的评分进行加权计算，得到各个维度的评分。
具体的，可以根据对各种驾驶特征设定的特征权值，将每个维度内的各项特征评分加权求和得到每个维度的评分，其中，每个维度内的特征权值之和为1。
维度评分计算公式为：维度评分＝∑特征评分*特征权值。
S315、对各个维度的评分进行加权计算，得到当前驾驶员的驾驶行为评分。
具体的，可以根据对各个维度设定的维度权值，对各个维度评分加权求和得到各个驾驶员的驾驶行为评分，其中，各维度权值之和为1。
驾驶行为评分计算公式为：驾驶行为评分＝∑维度评分*维度权值。
在一些实施例中，在上述参照图2描述的步骤S311～S315中，也可以不对当前驾驶员的驾驶特征进行维度划分，在获取该驾驶特征的评分之后，直接作为当前驾驶员的驾驶行为评分。
本公开实施例中，在当前阶段为冷启动阶段时，样本驾驶员的驾驶数据刚刚开始积累，并没有样本驾驶员的保险保单数据，通过使用冷启动阶段对应的驾驶评分方法对当前驾驶员进行驾驶行为评分，能够解决现有技术中由于系统运行初期样本数据积累不足而难以启动监督算法，进而导致无法进行驾驶行为评分的问题。
在本公开的一个实施例中，在当前阶段为样本数据的积累过程中的初期阶段时，积累了一定量的样本驾驶员的驾驶数据，样本驾驶员的保险保单数据仍然很少或者几乎没有，与驾驶员驾驶数据关联度整体比较低，可以使用基于特征分布函数的驾驶评分方法对当前驾驶员进行驾驶行为评分。具体的，如图3所示，前述步骤S3中使用当前积累阶段对应的驾驶评分方法对当前驾驶员的驾驶行为进行评分，可以包括步骤：
S321、从各样本驾驶员的样本驾驶数据中分别提取各样本驾驶员的多种样本驾驶特征。
具体的，该过程可参照步骤S311，此处不加以赘述。
S322、根据各样本驾驶员的一种或多种样本驾驶特征，计算各种样本驾驶特征的分布参数。
具体的，假设样本驾驶特征服从某种分布，可以对所有样本驾驶员的该样本驾驶特征进行汇总，并对汇总数据进行清洗，包括进行必要的异常值检测、数据归一化，基于清洗后的数据拟合得到该样本驾驶特征服从的分布函数的参数，以此类推，得到其他样本驾驶特征的分布参数。
示例性的，假如某一样本驾驶特征，如：平均每公里的急加速次数x服从指数分布，累积分布函数如下，其中λ acc>0，
Figure PCTCN2019095164-appb-000001
将所有样本驾驶员的该驾驶特征汇总，进行必要的异常值检测、数据归一化，基于清洗后的数据拟合得到该特征，也就是平均每公里的急加速次数服从的分布函数的参数λ acc。
S323、获取当前驾驶员的驾驶数据，并提取当前驾驶员的一种或多种驾驶特征。
具体的，该过程可参照步骤S311，此处不加以赘述。
S324、根据各种样本驾驶特征的分布参数，分别计算当前驾驶员的各种驾驶特征的累积分布函数，以作为各种驾驶特征对应的评分。
示例性的，对于当前驾驶员的某一驾驶特征，例如：平均每公里的急加速次数x obj，依据该驾驶特征服从的指数分布的参数λ acc得到该驾驶特征的累积分布函数值
Figure PCTCN2019095164-appb-000002
Figure PCTCN2019095164-appb-000003
以此作为该驾驶特征对应的评分。
S325、对各种驾驶特征的评分进行加权计算，得到当前驾驶员的驾驶行为评分。
本公开实施例中，在当前阶段为初期阶段时，积累了一定量的样本驾驶员的驾驶数据， 样本驾驶员的保险保单数据仍然很少或者几乎没有，与驾驶员驾驶数据关联度整体比较低，可以通过使用初期阶段对应的驾驶评分方法对当前驾驶员进行驾驶行为评分，能够解决现有技术中由于系统运行初期样本数据积累不足而难以启动监督算法，进而导致无法进行驾驶行为评分的问题。
在本公开的一个实施例中，在当前阶段为样本数据的积累过程中的中期阶段时，积累了一定量的样本驾驶员的驾驶数据与部分样本驾驶员的保险保单数据，可以使用基于聚类分析的驾驶评分方法对当前驾驶员进行驾驶行为评分。具体的，如图4所示，前述步骤S3中使用当前积累阶段对应的驾驶评分方法对当前驾驶员的驾驶行为进行评分，具体可以包括步骤：
S331、根据样本保险保单数据，获取与该样本保险保单数据对应的多个样本驾驶员的驾驶评分。
具体的，可以针对各样本驾驶员的保险保单数据，执行以下操作：
从保险保单数据中获取出险理赔数值，根据预设映射关系表，确定与出险理赔数值具有映射关系的驾驶评分，以作为样本驾驶员的驾驶评分；其中，在预设映射关系表中，理赔率越高对应的驾驶行为评分越低。
其中，出险理赔数值可以是从样本驾驶员的保险保单数据中提取出样本驾驶员的出险率或赔付率，根据出险率或理赔率由预设的映射关系表得到样本驾驶员的驾驶行为评分，以作为样本驾驶员的评分标签。其中，在映射关系表中，理赔率越高对应的驾驶行为评分越低，示例性地，分数范围可以为0～100。
S332、对所述多个样本驾驶员的驾驶数据进行聚类分析，获取多个聚类中心，每一类作为一个风险等级簇，并根据所述多个样本驾驶员的驾驶评分，计算各个风险等级簇的驾驶评分。
其中，可以使用K-means算法对部分样本驾驶员的驾驶数据进行聚类分析，每一类作为一个风险等级簇，获得一个聚类中心，认为同一类别中的驾驶员具有相同或者相似的风险水平，将同一类别中的样本驾驶员的驾驶评分的平均值作为该类别(即该风险等级簇)的驾驶评分。
S333、获取当前驾驶员的驾驶数据，并计算当前驾驶员的驾驶数据与多个聚类中心的相似度距离。
具体的，本公开实施例对具体的计算过程不加以限定。
S334、根据相似度距离的计算结果，确定当前驾驶员所对应的风险等级簇，并将确定出的风险等级簇的驾驶评分作为当前驾驶员的驾驶评分。
具体的，将相似度距离的计算结果中的最小值对应的风险等级簇作为当前驾驶员所对应的风险等级簇。
本公开实施例中，在当前阶段为样本数据的积累过程中的中期阶段时，积累了一定量的样本驾驶员的驾驶数据与部分样本驾驶员的保险保单数据，可以通过使用中期阶段对应的驾驶评分方法对当前驾驶员进行驾驶行为评分，能够解决现有技术中由于系统运行初期样本数据积累不足而难以启动监督算法，进而导致无法进行驾驶行为评分的问题。
在本公开的一个实施例中，在当前阶段为后期阶段时，样本驾驶员驾驶数据与理赔数据较为齐全，驾驶员驾驶数据与理赔数据关联比较完整，可以使用基于训练后的评分模型的驾驶评分方法对当前驾驶员进行驾驶行为评分。具体的，如图5所示，前述步骤S3中使用当前积累阶段对应的驾驶评分方法对当前驾驶员的驾驶行为进行评分，具体可以包括步骤：
S341、根据样本驾驶员的保险保单数据，获取与该样本保险保单数据对应的多个样本驾驶员的驾驶评分，并从样本驾驶数据中提取所述多个样本驾驶员的样本驾驶特征。
具体的，该步骤中，根据样本驾驶员的保险保单数据，获取与该样本保险保单数据对应的多个样本驾驶员的驾驶评分的过程与步骤S331相同，此处不再赘述。
该步骤中，从样本驾驶数据中提取所述多个样本驾驶员的样本驾驶特征的过程与步骤S311相同，此处不再赘述。
S342、根据各样本驾驶员的驾驶特征及各样本驾驶员的驾驶评分，建立驾驶行为评分模型。
具体的，对各样本驾驶员的驾驶特征及各样本驾驶员的驾驶评分，通过传统机器学习或者深度学习方法建立驾驶行为评分模型，并将驾驶行为评分模型进行离线存储，以供在线驾驶行为评分时进行调用。
其中，可以使用线性回归、随机森林、决策树、xgboost等方法建立驾驶行为评分模型，本公开对此不加以限定。
S343、使用驾驶行为评分模型对当前驾驶员的驾驶行为进行评分。
具体的，从当前驾驶员的驾驶数据中提取当前驾驶员的驾驶特征；将当前驾驶员的驾驶特征输入到驾驶行为评分模型中，得到当前驾驶员的驾驶行为评分并输出。
本公开实施例中，在当前阶段为样本数据的积累过程中的后期阶段时，由于样本驾驶员驾驶数据与理赔数据较为齐全，驾驶员驾驶数据与理赔数据关联比较完整，因此可以采用机器学习或深度学习训练得到驾驶评分模型，进而对当前驾驶员进行驾驶行为评分。
图6示出了根据本公开另一实施例的驾驶行为评分装置的示例性配置框图。本公开实 施例提供的驾驶行为评分装置可以用于执行上述实施例中的驾驶行为评分方法。
在一些实施例中，装置600可以包括处理电路60。装置600的处理电路60提供装置600的各种功能。在一些实施例中，装置600的处理电路60可以被配置为执行以上参照图1描述的驾驶行为评分方法。
处理电路60可以指在计算系统中执行功能的数字电路系统、模拟电路系统或混合信号(模拟和数字的组合)电路系统的各种实现。处理电路可以包括例如诸如集成电路(IC)、专用集成电路(ASIC)这样的电路、单独处理器核心的部分或电路、整个处理器核心、单独的处理器、诸如现场可编程门阵列(FPGA)的可编程硬件设备、和/或包括多个处理器的系统。
在一些实施例中，处理电路60可以包括积累模块61、确定模块62、评分模块63。
积累模块61被配置为积累样本数据，所述样本数据包括样本驾驶员的样本驾驶数据及样本保险保单数据；确定模块62被配置为根据所述样本驾驶数据及样本保险保单数据的积累量，确定所述样本数据的当前积累阶段；评分模块63被配置为使用所述当前积累阶段对应的驾驶评分方法，利用所积累的样本数据对当前驾驶员的驾驶行为进行评分。上述模块61～63可以分别被配置为执行前述图1中所示的驾驶行为评分方法中的步骤S1～步骤S3。
在一些实施例中，装置600还可以包括存储器(未图示)。装置600的存储器可以存储由处理电路60产生的信息以及用于装置600操作的程序和数据。存储器可以是易失性存储器和/或非易失性存储器。例如，存储器可以包括但不限于随机存取存储器(RAM)、动态随机存取存储器(DRAM)、静态随机存取存储器(SRAM)、只读存储器(ROM)以及闪存存储器。另外，装置600可以以芯片级来实现，或者也可以通过包括其它外部部件而以设备级来实现。
应当理解，上述各个模块仅是根据其所实现的具体功能所划分的逻辑模块，而不是用于限制具体的实现方式。在实际实现时，上述各个模块可被实现为独立的物理实体，或者也可由单个实体(例如，处理器(CPU或DSP等)、集成电路等)来实现。
本公开实施例提供的驾驶行为评分装置与本公开实施例所提供的驾驶行为评分方法属于同一发明构思，可执行本公开任意实施例所提供的驾驶行为评分方法，具备执行驾驶行为评分方法相应的功能模块和有益效果。未在本实施例中详尽描述的技术细节，可参见本公开实施例提供的驾驶行为评分方法，此处不再加以赘述。
图7示出了可以实现根据本公开的实施例的计算设备700的示例性配置。计算设备700是可以应用本公开的上述方面的硬件设备的实例。计算设备700可以是被配置为执行处理 和/或计算的任何机器。计算设备700可以是但不限制于工作站、服务器、台式计算机、膝上型计算机、平板计算机、个人数据助手(PDA)、智能电话、车载计算机或以上组合。
如图7所示，计算设备700可以包括可以经由一个或多个接口与总线702连接或通信的一个或多个元件。总线702可以包括但不限于，工业标准架构(Industry Standard Architecture，ISA)总线、微通道架构(Micro Channel Architecture，MCA)总线、增强ISA(EISA)总线、视频电子标准协会(VESA)局部总线、以及外设组件互连(PCI)总线等。计算设备700可以包括例如一个或多个处理器704、一个或多个输入设备706、以及一个或多个输出设备708。一个或多个处理器704可以是任何种类的处理器，并且可以包括但不限于一个或多个通用处理器或专用处理器(诸如专用处理芯片)。处理器704例如可以对应于图6中的处理电路60，被配置为实现本公开的校验证件与持证人的装置的各模块的功能。输入设备706可以是能够向计算设备输入信息的任何类型的输入设备，并且可以包括但不限于鼠标、键盘、触摸屏、麦克风和/或远程控制器。输出设备708可以是能够呈现信息的任何类型的设备，并且可以包括但不限于显示器、扬声器、视频/音频输出终端、振动器和/或打印机。
计算设备700还可以包括或被连接至非暂态存储设备714，该非暂态存储设备714可以是任何非暂态的并且可以实现数据存储的存储设备，并且可以包括但不限于盘驱动器、光存储设备、固态存储器、软盘、柔性盘、硬盘、磁带或任何其他磁性介质、压缩盘或任何其他光学介质、缓存存储器和/或任何其他存储芯片或模块、和/或计算机可以从其中读取数据、指令和/或代码的其他任何介质。计算设备700还可以包括随机存取存储器(RAM)710和只读存储器(ROM)712。ROM 712可以以非易失性方式存储待执行的程序、实用程序或进程。RAM 710可提供易失性数据存储，并存储与计算设备700的操作相关的指令。计算设备700还可包括耦接至数据链路718的网络/总线接口716。网络/总线接口716可以是能够启用与外部装置和/或网络通信的任何种类的设备或系统，并且可以包括但不限于调制解调器、网络卡、红外线通信设备、无线通信设备和/或芯片集(诸如蓝牙 TM设备、802.11设备、WiFi设备、WiMax设备、蜂窝通信设施等)。
此外，本公开另一实施例还提供了一种驾驶行为评分装置，包括：
一个或多个处理器；
存储装置，用于存储一个或多个程序；
当所述一个或多个程序被所述一个或多个处理器执行，使得所述一个或多个处理器实现如上述实施例所述的驾驶行为评分方法。
此外，本公开另一实施例还提供了一种计算机可读存储介质，其上存储有计算机程序， 所述程序被处理器执行时实现如上述实施例所述的驾驶行为评分方法。
需要说明的是，在本公开的描述中，术语“第一”、“第二”等仅用于描述目的，而不能理解为指示或暗示相对重要性和顺序。
本领域内的技术人员应明白，本公开实施例中的实施例可提供为方法、系统、或计算机程序产品。因此，本公开实施例中可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本公开实施例中可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本公开实施例中是参照根据本公开实施例中实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
尽管已描述了本公开实施例中的优选实施例，但本领域内的技术人员一旦得知了基本创造性概念，则可对这些实施例作出另外的变更和修改。所以，所附权利要求意欲解释为包括优选实施例以及落入本公开实施例中范围的所有变更和修改。
显然，本领域的技术人员可以对本公开进行各种改动和变型而不脱离本公开的精神和范围。这样，倘若本公开的这些修改和变型属于本公开权利要求及其等同技术的范围之内，则本公开也意图包含这些改动和变型在内。

Claims (18)

  1. 一种驾驶行为评分方法，其特征在于，所述方法包括步骤：
    积累样本数据，所述样本数据包括样本驾驶员的样本驾驶数据及样本保险保单数据；
    根据所述样本驾驶数据及样本保险保单数据的积累量，确定所述样本数据的当前积累阶段；
    使用所述当前积累阶段对应的驾驶评分方法，利用所积累的样本数据对当前驾驶员的驾驶行为进行评分。
  2. 根据权利要求1所述的方法，其特征在于，
    在所述样本保险保单数据的积累量低于第一阈值时，确定所述样本数据的当前积累阶段为第一阶段，
    使用所述第一阶段对应的驾驶评分方法，利用所积累的样本数据中的样本驾驶数据对所述当前驾驶员的驾驶行为进行评分。
  3. 根据权利要求1所述的方法，其特征在于，
    在所述样本保险保单数据的积累量在第一阈值以上时，确定所述样本数据的当前积累阶段为第二阶段，
    使用所述第二阶段对应的驾驶评分方法，利用所积累的样本数据中的样本驾驶数据及样本保险保单数据对所述当前驾驶员的驾驶行为进行评分。
  4. 根据权利要求2所述的方法，其特征在于，
    所述第一阶段包括所述样本保险保单数据的积累量低于所述第一阈值并且所述样本驾驶数据的积累量低于第二阈值的第三阶段，
    在所述当前积累阶段为所述第三阶段时，所述对当前驾驶员的驾驶行为进行评分包括：
    获取所述当前驾驶员的驾驶数据，并提取所述当前驾驶员的驾驶特征；
    根据预设评分标准，获取所述当前驾驶员的驾驶特征的评分，作为所述当前驾驶员的驾驶行为评分。
  5. 根据权利要求2所述的方法，其特征在于，
    所述第一阶段包括所述样本保险保单数据的积累量低于所述第一阈值并且所述样本驾驶数据的积累量在第二阈值以上的第四阶段，
    在所述当前积累阶段为所述第四阶段时，所述对当前驾驶员的驾驶行为进行评分包括：
    从所述样本驾驶员的样本驾驶数据中提取所述样本驾驶员的样本驾驶特征；
    根据所述样本驾驶员的样本驾驶特征，计算所述样本驾驶特征的分布参数；
    获取所述当前驾驶员的驾驶数据，并提取所述当前驾驶员的驾驶特征；
    根据所述样本驾驶特征的分布参数，计算所述当前驾驶员的驾驶特征的累积分布函数，作为所述当前驾驶员的驾驶行为评分。
  6. 根据权利要求3所述的方法，其特征在于，
    所述第二阶段包括所述样本保险保单数据的积累量在所述第一阈值以上、并且所述样本保险保单数据的积累量小于所述样本驾驶数据的积累量的第五阶段，
    在所述当前积累阶段为所述第五阶段时，所述对当前驾驶员的驾驶行为进行评分包括：
    根据所述样本保险保单数据，获取与该样本保险保单数据对应的多个样本驾驶员的驾驶评分；
    对所述多个样本驾驶员的驾驶数据进行聚类分析，获取多个聚类中心，每一类作为一个风险等级簇，并根据所述多个样本驾驶员的驾驶评分，计算各个风险等级簇的驾驶评分；
    获取所述当前驾驶员的驾驶数据，并计算所述当前驾驶员的驾驶数据与所述多个聚类中心的相似度距离；
    根据所述相似度距离的计算结果，确定所述当前驾驶员所对应的风险等级簇，并将确定出的所述风险等级簇的驾驶评分作为所述当前驾驶员的驾驶评分。
  7. 根据权利要求3所述的方法，其特征在于，
    所述第二阶段包括所述样本保险保单数据的积累量在所述第一阈值以上、并且所述样本保险保单数据的积累量等于所述样本驾驶数据的积累量的第六阶段，
    在所述当前积累阶段为所述第六阶段时，所述对当前驾驶员的驾驶行为进行评分包括：
    根据所述样本保险保单数据，获取与该样本保险保单数据对应的多个样本驾驶员的驾驶评分；
    从所述样本驾驶数据中提取所述多个样本驾驶员的样本驾驶特征；
    根据所述多个样本驾驶员的样本驾驶特征及对应的驾驶评分，建立驾驶行为评分模型；
    使用所述驾驶行为评分模型对所述当前驾驶员的驾驶行为进行评分。
  8. 根据权利要求6或7所述的方法，其特征在于，所述根据所述样本保险保单数据，获取与该保单数据对应的多个样本驾驶员的驾驶评分包括：
    针对各所述多个样本驾驶员的样本保险保单数据，执行以下操作：
    从所述样本保险保单数据中获取出险理赔数值；
    根据预设映射关系表，确定与所述出险理赔数值具有映射关系的驾驶评分，以作为所述样本驾驶员的驾驶评分。
  9. 一种驾驶行为评分装置，其特征在于，所述装置包括：
    积累模块，被配置为积累样本数据，所述样本数据包括样本驾驶员的样本驾驶数据及样本保险保单数据；
    确定模块，被配置为根据所述样本驾驶数据及样本保险保单数据的积累量，确定所述样本数据的当前积累阶段；
    评分模块，被配置为使用所述当前积累阶段对应的驾驶评分方法，利用所积累的样本数据对当前驾驶员的驾驶行为进行评分。
  10. 根据权利要求9所述的装置，其特征在于，
    在所述样本保险保单数据的积累量低于第一阈值时，确定所述样本数据的当前积累阶段为第一阶段，
    使用所述第一阶段对应的驾驶评分方法，利用所积累的样本数据中的样本驾驶数据对所述当前驾驶员的驾驶行为进行评分。
  11. 根据权利要求9所述的装置，其特征在于，
    在所述样本保险保单数据的积累量在第一阈值以上时，确定所述样本数据的当前积累阶段为第二阶段，
    使用所述第二阶段对应的驾驶评分方法，利用所积累的样本数据中的样本驾驶数据及样本保险保单数据对所述当前驾驶员的驾驶行为进行评分。
  12. 根据权利要求10所述的装置，其特征在于，
    所述第一阶段包括所述样本保险保单数据的积累量低于所述第一阈值并且所述样本驾驶数据的积累量低于第二阈值的第三阶段，
    在所述当前积累阶段为所述第三阶段时，所述评分模块被配置为：
    获取所述当前驾驶员的驾驶数据，并提取所述当前驾驶员的驾驶特征；
    根据预设评分标准，获取所述当前驾驶员的驾驶特征的评分，作为所述当前驾驶员的驾驶行为评分。
  13. 根据权利要求10所述的装置，其特征在于，
    所述第一阶段包括所述样本保险保单数据的积累量低于所述第一阈值并且所述样本驾驶数据的积累量在第二阈值以上的第四阶段，
    在所述当前积累阶段为所述第四阶段时，所述评分模块被配置为：
    从所述样本驾驶员的样本驾驶数据中提取所述样本驾驶员的样本驾驶特征；
    根据所述样本驾驶员的样本驾驶特征，计算所述样本驾驶特征的分布参数；
    获取所述当前驾驶员的驾驶数据，并提取所述当前驾驶员的驾驶特征；
    根据所述样本驾驶特征的分布参数，计算所述当前驾驶员的驾驶特征的累积分布函数， 作为所述当前驾驶员的驾驶行为评分。
  14. 根据权利要求11所述的装置，其特征在于，
    所述第二阶段包括所述样本保险保单数据的积累量在所述第一阈值以上、并且所述样本保险保单数据的积累量小于所述样本驾驶数据的积累量的第五阶段，
    在所述当前积累阶段为所述第五阶段时，所述评分模块被配置为：
    根据所述样本保险保单数据，获取与该样本保险保单数据对应的多个样本驾驶员的驾驶评分；
    对所述多个样本驾驶员的驾驶数据进行聚类分析，获取多个聚类中心，每一类作为一个风险等级簇，并根据所述多个样本驾驶员的驾驶评分，计算各个风险等级簇的驾驶评分；
    获取所述当前驾驶员的驾驶数据，并计算所述当前驾驶员的驾驶数据与所述多个聚类中心的相似度距离；
    根据所述相似度距离的计算结果，确定所述当前驾驶员所对应的风险等级簇，并将确定出的所述风险等级簇的驾驶评分作为所述当前驾驶员的驾驶评分。
  15. 根据权利要求11所述的装置，其特征在于，
    所述第二阶段包括所述样本保险保单数据的积累量在所述第一阈值以上、并且所述样本保险保单数据的积累量等于所述样本驾驶数据的积累量的第六阶段，
    在所述当前积累阶段为所述第六阶段时，所述评分模块被配置为：
    根据所述样本保险保单数据，获取与该样本保险保单数据对应的多个样本驾驶员的驾驶评分；
    从所述样本驾驶数据中提取所述多个样本驾驶员的样本驾驶特征；
    根据所述多个样本驾驶员的样本驾驶特征及对应的驾驶评分，建立驾驶行为评分模型；
    使用所述驾驶行为评分模型对所述当前驾驶员的驾驶行为进行评分。
  16. 根据权利要求14或15所述的装置，其特征在于，所述根据所述样本保险保单数据，获取与该保单数据对应的多个样本驾驶员的驾驶评分包括：
    针对各所述多个样本驾驶员的样本保险保单数据，执行以下操作：
    从所述样本保险保单数据中获取出险理赔数值；
    根据预设映射关系表，确定与所述出险理赔数值具有映射关系的驾驶评分，以作为所述样本驾驶员的驾驶评分。
  17. 一种驾驶行为评分装置，其特征在于，所述装置包括：
    一个或多个处理器；
    存储装置，用于存储一个或多个程序；
    当所述一个或多个程序被所述一个或多个处理器执行，使得所述一个或多个处理器实现如权利要求1～8任意一项所述的驾驶行为评分方法。
  18. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，所述程序被处理器执行时实现如权利要求1～8任意一项所述的驾驶行为评分方法。
PCT/CN2019/095164 2018-11-28 2019-07-09 一种驾驶行为评分方法、装置及计算机可读存储介质 WO2020107894A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
JP2019554753A JP6918137B2 (ja) 2018-11-28 2019-07-09 運転行為評価方法、装置及びコンピュータが読み取り可能な記憶媒体
SG11202005593WA SG11202005593WA (en) 2018-11-28 2019-07-09 Driving behavior scoring method, device and computer readable storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811433626.6 2018-11-28
CN201811433626.6A CN109670970B (zh) 2018-11-28 2018-11-28 一种驾驶行为评分方法、装置及计算机可读存储介质

Publications (1)

Publication Number Publication Date
WO2020107894A1 true WO2020107894A1 (zh) 2020-06-04

Family

ID=66143248

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/095164 WO2020107894A1 (zh) 2018-11-28 2019-07-09 一种驾驶行为评分方法、装置及计算机可读存储介质

Country Status (4)

Country Link
JP (1) JP6918137B2 (zh)
CN (1) CN109670970B (zh)
SG (1) SG11202005593WA (zh)
WO (1) WO2020107894A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113298361A (zh) * 2021-05-06 2021-08-24 深圳市锐明技术股份有限公司 一种危险驾驶行为的评价方法、装置、电子设备及系统
CN115953858A (zh) * 2022-11-29 2023-04-11 摩尔线程智能科技(北京)有限责任公司 一种基于车载dms的驾驶评分方法、装置及电子设备

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109670970B (zh) * 2018-11-28 2020-12-22 众安信息技术服务有限公司 一种驾驶行为评分方法、装置及计算机可读存储介质
CN112749721A (zh) * 2019-10-31 2021-05-04 彩虹无线（北京）新技术有限公司 驾驶风险评价模型训练方法及装置
CN112937591B (zh) * 2019-12-11 2023-04-07 彩虹无线（北京）新技术有限公司 驾驶安全监测方法、装置、设备及计算机可读存储介质
CN113119981B (zh) * 2021-04-09 2022-06-17 东风汽车集团股份有限公司 车辆主动安全控制方法、系统及存储介质
CN113635915B (zh) * 2021-08-24 2023-01-06 中国人民解放军陆军装甲兵学院 车辆驾驶预警方法、装置、电子设备和存储介质
CN114407904A (zh) * 2022-02-11 2022-04-29 西北工业大学 驾驶员侵略性驾驶行为监测方法、装置及系统
CN116653980B (zh) * 2023-06-28 2024-08-06 运脉云技术有限公司 一种驾驶员驾驶习惯分析系统及驾驶习惯分析方法
CN116821805A (zh) * 2023-06-28 2023-09-29 运脉云技术有限公司 一种驾驶行为监控的车务平台系统及驾驶行为监控方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106127586A (zh) * 2016-06-17 2016-11-16 上海经达信息科技股份有限公司 大数据时代下车险费率辅助决策系统
CN107203943A (zh) * 2017-04-06 2017-09-26 北京保程保险公估有限公司 机动车商业保险定价系统
CN107845039A (zh) * 2016-09-20 2018-03-27 得道车联网络科技（上海）有限公司 一种无尺度自适应的车联网车险评分模型
CN108492053A (zh) * 2018-04-11 2018-09-04 北京汽车研究总院有限公司 驾驶员风险评估模型训练、风险评估方法和装置
WO2018180347A1 (ja) * 2017-03-29 2018-10-04 ソニー株式会社 情報処理装置、情報処理システム、および情報処理方法、並びにプログラム
CN109670970A (zh) * 2018-11-28 2019-04-23 众安信息技术服务有限公司 一种驾驶行为评分方法、装置及计算机可读存储介质

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4978747B2 (ja) * 2011-11-16 2012-07-18 日産自動車株式会社 車両用運転支援装置および車両用運転支援方法
CN107784251A (zh) * 2016-08-25 2018-03-09 大连楼兰科技股份有限公司 基于图像识别技术对驾驶行为评价的方法
CN107784587B (zh) * 2016-08-25 2021-09-14 大连楼兰科技股份有限公司 一种驾驶行为评价系统
CN107038860B (zh) * 2016-11-18 2019-07-23 杭州好好开车科技有限公司 一种基于adas技术及回归模型的用户驾驶行为评分方法
CN106651133B (zh) * 2016-11-18 2020-07-10 杭州好好开车科技有限公司 一种使用手机实现基于adas的用户驾驶行为评分方法
CN106815312A (zh) * 2016-12-21 2017-06-09 东软集团股份有限公司 一种驾驶员评价方法及装置
JP7074069B2 (ja) * 2016-12-22 2022-05-24 ソニーグループ株式会社 情報処理装置および方法、並びにプログラム
CN107203945A (zh) * 2017-06-12 2017-09-26 清华大学苏州汽车研究院（吴江） 车险分级评估方法及装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106127586A (zh) * 2016-06-17 2016-11-16 上海经达信息科技股份有限公司 大数据时代下车险费率辅助决策系统
CN107845039A (zh) * 2016-09-20 2018-03-27 得道车联网络科技（上海）有限公司 一种无尺度自适应的车联网车险评分模型
WO2018180347A1 (ja) * 2017-03-29 2018-10-04 ソニー株式会社 情報処理装置、情報処理システム、および情報処理方法、並びにプログラム
CN107203943A (zh) * 2017-04-06 2017-09-26 北京保程保险公估有限公司 机动车商业保险定价系统
CN108492053A (zh) * 2018-04-11 2018-09-04 北京汽车研究总院有限公司 驾驶员风险评估模型训练、风险评估方法和装置
CN109670970A (zh) * 2018-11-28 2019-04-23 众安信息技术服务有限公司 一种驾驶行为评分方法、装置及计算机可读存储介质

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113298361A (zh) * 2021-05-06 2021-08-24 深圳市锐明技术股份有限公司 一种危险驾驶行为的评价方法、装置、电子设备及系统
CN115953858A (zh) * 2022-11-29 2023-04-11 摩尔线程智能科技(北京)有限责任公司 一种基于车载dms的驾驶评分方法、装置及电子设备

Also Published As

Publication number Publication date
CN109670970A (zh) 2019-04-23
SG11202005593WA (en) 2020-07-29
JP2021509978A (ja) 2021-04-08
CN109670970B (zh) 2020-12-22
JP6918137B2 (ja) 2021-08-11

Similar Documents

Publication Publication Date Title
WO2020107894A1 (zh) 一种驾驶行为评分方法、装置及计算机可读存储介质
CN109118055B (zh) 一种驾驶行为评分方法及装置
Papadimitriou et al. Analysis of driver behaviour through smartphone data: The case of mobile phone use while driving
US11375338B2 (en) Method for smartphone-based accident detection
US11216888B2 (en) Electronic system for dynamic, quasi-realtime measuring and identifying driver maneuvers solely based on mobile phone telemetry, and a corresponding method thereof
KR101617349B1 (ko) 운전 습관 분석 진단 시스템 및 방법
Osafune et al. Analysis of accident risks from driving behaviors
JP7398383B2 (ja) テレマティクスデータに基づく車両分類
US10990837B1 (en) Systems and methods for utilizing machine learning and feature selection to classify driving behavior
US10899295B2 (en) Using data collected by a personal electronic device to identify a vehicle
AU2021202991B2 (en) Method and system for vehicle speed profile generation
US12038294B2 (en) Machine learning platform for dynamic device and sensor quality evaluation
CN106651133B (zh) 一种使用手机实现基于adas的用户驾驶行为评分方法
US12106617B2 (en) Method and system for auto generating automotive data quality marker
CN115841712A (zh) 一种基于v2x技术的驾驶数据处理方法、装置及设备
JP7552727B2 (ja) 運転評価システム、学習装置、評価結果出力装置、方法およびプログラム
CN115619572A (zh) 车险报价方法、装置、设备及存储介质
Kontaxi et al. Discovering the influence of feedback on driver behavior through a multiphase experiment based on a smartphone application
Malik et al. Enhancing transportation safety: An integrated approach using FLFS and OSNCA for advanced driving behavior analysis
US11934364B2 (en) System and method facilitating determination of automotive signal quality marker
CN118607932B (zh) 基于深度学习的驾驶人个性化风险评价方法、电子设备和存储介质
CN117787699A (zh) 一种道路风险预测方法、装置、计算机设备及存储介质
CN115626169A (zh) 基于gps行程数据评估驾驶路线熟悉度的方法及设备
Hosseinioun Employing sensor and service fusion to assess driving performance
CN115392515A (zh) 一种驾驶行为数据处理方法、计算机设备及计算机存储介质

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2019554753

Country of ref document: JP

Kind code of ref document: A

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

Ref document number: 19888723

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 09.09.2021)

122 Ep: pct application non-entry in european phase

Ref document number: 19888723

Country of ref document: EP

Kind code of ref document: A1