JP2021509978A - Driving behavior evaluation method, device and computer-readable storage medium - Google Patents

Abstract

The present disclosure relates to driving behavior evaluation methods, devices and computer-readable storage media. The driving behavior evaluation method is to accumulate sample data including sample driving data and sample insurance contract data of a sample driver, and based on the accumulated amount of the sample driving data and sample insurance contract data, the current accumulation step of the sample data. Includes determining the current driver's driving behavior using the accumulated sample data using the driving evaluation method corresponding to the currently accumulating staircase.

Description

The present disclosure relates to the technical field of vehicle driving behavior analysis, and particularly to driving behavior evaluation methods, devices and computer-readable storage media.

In recent years, the number of automobiles has increased, and at the same time, with the rapid development of car networking, more and more automobile insurance products have a fixed price based on the driver's driving behavior combined with car networking technology.

UBI (Usage Based Insurance) automobile insurance is a vehicle insurance that sets a fixed price based on driving behavior and vehicle usage data, and is a new automobile insurance that has emerged under the background of the liberalization of automobile insurance premium rates. By collecting driving behavior data and vehicle driving data, the insurance company estimates the driving behavior risk level of the driver through the data analysis process. For car owners who drive well, the insurance company can give them a lot of discounts, and for owners of cars who drive badly, the insurance company can increase their premiums accordingly. This can help car owners develop good driving practices, reduce accidents, reduce risks and complaints, improve traffic conditions and reduce social costs.

Therefore, for insurance companies, before setting a fixed price for automobile insurance products, first estimate the driving behavior risk based on the driving data of the driver, and now always adopt the method of supervision learning in device learning. It is necessary to acquire the relationship between driving behavior and driving risk, and to include driving behavior and vehicle usage data and insurance claim data at the same time in the training sample. In collecting driving behavior and vehicle usage data, always use on-board OBD (On-Board Diagnostics) equipment or equipment such as smartphones to acquire vehicle driving data with sensors in the equipment, and if the equipment has a camera, drive with the camera. By acquiring the video, you can also obtain more data that embodies driving habits. As a correct ruler of risk, it is necessary to obtain a situation in which the observed driver has had a traffic accident within the same cycle of observing the driving behavior and demands compensation from the insurance company, and the risk probability or compensation rate is determined. It is a sign of driving risk. After acquiring both of the above, the driving data is used as a feature in the supervision algorithm, the risk probability or the compensation rate is used as the supervision signal in the supervision algorithm, and the supervision algorithm is based on the training data and the relationship between the feature and the supervision signal, that is, driving. Obtain the relationship between data, risk probability, and damage compensation rate. For a target sample for which it is necessary to determine the driving behavior risk, that is, a driver who has only driving data, the driving behavior risk is acquired based on the inference relationship obtained by the supervision algorithm.

Hereinafter, the present disclosure will be briefly presented to provide a basic understanding of some of the technical proposals of the present disclosure. However, it should be understood that this overview is not an exhaustive overview of this disclosure. It is not intended to determine the main or important parts of this disclosure, nor is it intended to limit the scope of this disclosure. Its purpose is simply to present some notions of this disclosure in a simple form, which will be used as a prelude to the detailed description described below.

In the embodiment of the present disclosure, a driving behavior evaluation method, a device, and a computer-readable memory for accurately predicting a driving behavior evaluation for a target driver by using different driving evaluation methods on stairs with different data accumulation. The purpose is to provide a medium. Specific technical proposals for the examples of the present disclosure are as follows.

According to the first direction, the embodiment of the present disclosure is a driving behavior evaluation method.
Accumulating sample data including sample driver data and sample insurance contract data,
To determine the current accumulation staircase of the sample data based on the accumulated amount of the sample operation data and the sample insurance contract data.
Provided is a driving behavior evaluation method including using the driving evaluation method corresponding to the currently accumulated stairs and evaluating the driving behavior of the current driver using the accumulated sample data.

In some embodiments, when the accumulated amount of the sample insurance contract data is lower than the first threshold value, the currently accumulated staircase of the sample data is determined as the first staircase.
The driving evaluation method corresponding to the first staircase is used, and the driving behavior of the current driver is evaluated by using the sample driving data in the accumulated sample data.

In some embodiments, when the accumulated amount of the sample insurance contract data is equal to or greater than the first threshold value, the currently accumulated staircase of the sample data is determined as the second staircase.
The driving evaluation method corresponding to the second staircase is used, and the driving behavior of the current driver is evaluated by using the sample driving data and the sample insurance contract data in the accumulated sample data.

In some embodiments, the first staircase is a third staircase in which the accumulated amount of sample insurance contract data is lower than the first threshold value and the accumulated amount of sample operation data is lower than the second threshold value. Including
When the present accumulation staircase is the third staircase, assessing the driving behavior of the current driver
To acquire the driving data of the current driver and extract the driving characteristics of the current driver.
This includes acquiring the evaluation of the driving characteristics of the current driver as the driving behavior evaluation of the current driver based on the preliminary evaluation standard.

In some embodiments, the first staircase is a fourth staircase in which the accumulated amount of sample insurance contract data is lower than the first threshold value and the accumulated amount of sample operation data is equal to or greater than the second threshold value. Including
When the present accumulation staircase is the fourth staircase, assessing the driving behavior of the current driver
Extracting the sample driving characteristics of the sample driver from the sample driving data of the sample driver,
To calculate the distribution parameters of the sample driving feature based on the sample driving feature of the sample driver.
To acquire the driving data of the current driver and extract the driving characteristics of the current driver.
It includes calculating the cumulative distribution function of the current driver's driving characteristics as the driving behavior evaluation of the current driver based on the distribution parameters of the sample driving characteristics.

In some embodiments, in the second staircase, the accumulated amount of the sample insurance contract data is equal to or greater than the first threshold value, and the accumulated amount of the sample insurance contract data is larger than the accumulated amount of the sample operation data. Including the lower fifth staircase
When the present accumulation staircase is the fifth staircase, assessing the driving behavior of the current driver
Obtaining driving evaluations of a plurality of sample drivers corresponding to the sample insurance contract data based on the sample insurance contract data.
The driving data of the plurality of sample drivers is clustered and analyzed, the centers of the plurality of clusters are acquired, each cluster is made into one risk level cluster, and each risk level cluster is based on the driving evaluation of the plurality of sample drivers. To calculate the driving evaluation of
To acquire the driving data of the current driver and calculate the similarity distance between the driving data of the current driver and the centers of the plurality of clusters.
This includes determining the risk level cluster corresponding to the current driver based on the calculation result of the similarity distance, and setting the determined driving evaluation of the risk level cluster as the driving evaluation of the current driver.

In some embodiments, in the second staircase, the accumulated amount of the sample insurance contract data is equal to or greater than the first threshold value, and the accumulated amount of the sample insurance contract data is the accumulated amount of the sample operation data. Including the equal sixth staircase
When the present accumulation staircase is the sixth staircase, assessing the driving behavior of the current driver
Obtaining driving evaluations of a plurality of sample drivers corresponding to the sample insurance contract data based on the sample insurance contract data.
Extracting sample driving features of the plurality of sample drivers from the sample driving data,
Building a driving behavior evaluation model based on the sample driving characteristics of the plurality of sample drivers and the corresponding driving evaluations.
This includes evaluating the current driver's driving behavior using the driving behavior evaluation model.

In some embodiments, it is possible to obtain a driving evaluation of a plurality of sample drivers corresponding to the contract data based on the sample insurance contract data described above.
For sample insurance contract data of each of the plurality of sample drivers
Obtaining insurance claim figures from the sample insurance contract data,
Based on the preliminary mapping relationship table, the driving evaluation having a mapping relationship with the insurance claim value is determined as the driving evaluation of the sample driver.

According to the second direction, the embodiment of the present disclosure is a driving behavior evaluation device.
A storage module arranged to store sample data, including sample driver data and sample insurance contract data, and a storage module.
Based on the accumulated amount of the sample operation data and the sample insurance contract data, the confirmation module arranged so as to determine the current accumulation step of the sample data, and
Provided is a driving behavior evaluation device including an evaluation module arranged to evaluate the driving behavior of the current driver using the accumulated sample data by using the driving evaluation method corresponding to the currently accumulated staircase. ..

In some embodiments, when the accumulated amount of the sample insurance contract data is lower than the first threshold value, the currently accumulated staircase of the sample data is determined as the first staircase.
The driving evaluation method corresponding to the first staircase is used, and the driving behavior of the current driver is evaluated by using the sample driving data in the accumulated sample data.

In some embodiments, when the accumulated amount of the sample insurance contract data is equal to or greater than the first threshold value, the currently accumulated staircase of the sample data is determined as the second staircase.
The driving evaluation method corresponding to the second staircase is used, and the driving behavior of the current driver is evaluated by using the sample driving data and the sample insurance contract data in the accumulated sample data.

In some embodiments, the first staircase is a third staircase in which the accumulated amount of sample insurance contract data is lower than the first threshold value and the accumulated amount of sample operation data is lower than the second threshold value. Including
When the current storage staircase is the third staircase, the evaluation module
The driving data of the current driver is acquired, and the driving characteristics of the current driver are extracted.
Based on the preliminary evaluation standard, the evaluation of the driving characteristics of the current driver is arranged so as to be acquired as the driving behavior evaluation of the current driver.

In some embodiments, the first staircase is a fourth staircase in which the accumulated amount of sample insurance contract data is lower than the first threshold value and the accumulated amount of sample operation data is equal to or greater than the second threshold value. Including
When the current storage staircase is the fourth staircase, the evaluation module
From the sample driving data of the sample driver, the sample driving characteristics of the sample driver are extracted.
Based on the sample driving characteristics of the sample driver, the distribution parameters of the sample driving characteristics are calculated.
The driving data of the current driver is acquired, and the driving characteristics of the current driver are extracted.
Based on the distribution parameter of the sample driving feature, the cumulative distribution function of the driving feature of the current driver is arranged so as to be calculated as a driving behavior evaluation of the current driver.

In some embodiments, in the second staircase, the accumulated amount of the sample insurance contract data is equal to or greater than the first threshold value, and the accumulated amount of the sample insurance contract data is larger than the accumulated amount of the sample operation data. Including the lower fifth staircase
When the current storage staircase is the fifth staircase, the evaluation module
Based on the sample insurance contract data, the driving evaluations of a plurality of sample drivers corresponding to the sample insurance contract data are acquired.
The driving data of the plurality of sample drivers is clustered and analyzed, the centers of the plurality of clusters are acquired, each cluster is made into one risk level cluster, and each risk level cluster is based on the driving evaluation of the plurality of sample drivers. Calculate the driving evaluation of
The driving data of the current driver is acquired, and the similarity distance between the driving data of the current driver and the centers of the plurality of clusters is calculated.
Based on the calculation result of the similarity distance, the risk level cluster corresponding to the current driver is determined, and the determined driving evaluation of the risk level cluster is arranged so as to be the driving evaluation of the current driver. There is.

In some embodiments, in the second staircase, the accumulated amount of the sample insurance contract data is equal to or greater than the first threshold value, and the accumulated amount of the sample insurance contract data is the accumulated amount of the sample operation data. Including the equal sixth staircase
When the current storage staircase is the sixth staircase, the evaluation module
Based on the sample insurance contract data, the driving evaluations of a plurality of sample drivers corresponding to the sample insurance contract data are acquired.
From the sample driving data, sample driving features of the plurality of sample drivers are extracted.
Based on the sample driving characteristics of the plurality of sample drivers and the corresponding driving evaluations, a driving behavior evaluation model was constructed.
It is arranged to evaluate the current driver's driving behavior using the driving behavior evaluation model.

In some embodiments, it is possible to obtain a driving evaluation of a plurality of sample drivers corresponding to the contract data based on the sample insurance contract data described above.
For sample insurance contract data of each of the plurality of sample drivers
Obtaining insurance claim figures from the sample insurance contract data,
Based on the preliminary mapping relationship table, the driving evaluation having a mapping relationship with the insurance claim value is determined as the driving evaluation of the sample driver.

According to the third direction, the embodiment of the present disclosure is a driving behavior evaluation device.
With one or more processors
Includes a storage device for storing one or more programs,
When the one or more programs are executed by the one or more processors, a driving behavior evaluation device that causes the one or more processors to realize the driving behavior evaluation method described in the first technical proposal is provided. provide.

According to a fourth aspect, an embodiment of the present disclosure is a computer-readable storage medium that stores a computer program and is described in the first technical proposal when the program is executed by a processor. Provide a computer-readable storage medium that realizes a driving behavior evaluation method.

The driving behavior evaluation method, device, and computer-readable storage medium provided in the embodiments of the present disclosure accumulate sample data, determine the current accumulation staircase of sample data based on the amount of sample accumulation, and present storage staircase. Evaluate the current driver's driving behavior using the corresponding driving evaluation method. This makes it possible to estimate the driving behavior risk at any of the stairs with different data storage.

In order to more clearly explain the technical proposals in the examples of the present disclosure, the drawings used in the description of the examples will be briefly introduced below, and clearly, the drawings described below are some of the drawings of the present disclosure. It is only an embodiment, and for those skilled in the art, other drawings can be obtained based on those drawings on the premise that they do not do creative work.
FIG. 1 shows a typical flow chart of a driving behavior evaluation method according to an embodiment of the present disclosure. FIG. 2 shows a typical flow chart of the driving behavior evaluation method according to another embodiment of the present disclosure. FIG. 3 shows a typical flow chart of the driving behavior evaluation method according to another embodiment of the present disclosure. FIG. 4 shows a typical flow chart of the driving behavior evaluation method according to another embodiment of the present disclosure. FIG. 5 shows a typical flow chart of the driving behavior evaluation method according to another embodiment of the present disclosure. FIG. 6 shows a typical layout block diagram of a driving behavior evaluation device according to another embodiment of the present disclosure. FIG. 7 shows a typical arrangement of computing equipment that can realize the embodiments of the present disclosure.

In order to further clarify the object, technical proposal and advantages of the present invention, the technical proposal in the examples of the present disclosure shall be clearly and completely described and clearly described below with reference to the drawings in the examples of the present disclosure. The examples described are only partial examples of the present disclosure, not all examples. According to the examples in this disclosure, all other examples obtained under the premise that those skilled in the art do not do creative work are all within the scope of the protection of this disclosure.

When estimating the driving behavior risk based on the driving data of the driver, one method that the inventor of the present disclosure knows is to adopt the method of supervised learning in machine learning to determine the relationship between the driving behavior and the driving risk. It is something to acquire. However, this kind of supervised learning method needs to accumulate a large amount of driving data and insurance claim data, and the insurance claim cycle needs to match the collection cycle of driving data. Without insurance claim data, or without a certain number of data, the algorithm is difficult to invoke. In actual operation, collecting driving data is often relatively convenient in the early stages of system operation, but risks and complaints do not always occur immediately, so collecting driving data and at the same time acquiring insurance claim data. Can't. At the same time, according to the driving itinerary and the accumulation of data, even if a danger occurs, the damage compensation flow includes multiple steps such as investigation, accident fixed loss, repair, damage compensation, etc., so a certain amount of time is required, and it is a risk. There may be a certain amount of time delay in retrieving claim data, and the supervision algorithm is also difficult to activate.

The embodiments of the present disclosure provide a driving behavior evaluation method capable of accurately predicting a driving behavior evaluation for a target driver by using different driving behavior evaluation methods on stairs with different data accumulation.

The execution subject of the driving behavior evaluation method can be a server. The server can make a communication connection with a mobile terminal equipped with a mobile SDK (Soft Development Kit) module and an in-vehicle deep camera module via a network. The mobile SDK module can be, for example, a software development kit provided by a mobile terminal, the disclosure of which is not limited thereto.

The server may connect to the insurance contract system through a preliminary interface to acquire the driver's insurance contract data. The preliminary interface for the present disclosure is some predefined functions in the software system for interacting, data transmission, etc. with other systems, and the present disclosure does not limit the preliminary interface. ..

The server may be a single server, a server group composed of a plurality of servers, and a communication connection can be made between a plurality of servers within the server group. The mobile terminal equipped with the mobile SDK module can be the driver's mobile terminal, which can collect the driver's driving behavior data by the mobile SDK module and upload it to the server. The in-vehicle deep camera module can be attached to the driver's vehicle and can upload the collected driving video data to the server.

FIG. 1 shows a typical flow chart of a driving behavior evaluation method according to an embodiment of the present disclosure.

As shown in FIG. 1, in step S1, sample driving data including sample driving data and sample insurance contract data of a sample driver are accumulated.

In some embodiments, the sample driving data of the sample driver can include at least one of driving behavior data, driving environment data and driving video data.

In some embodiments, the mobile SDK module can acquire driving behavior data for the sample driver. Driving behavior data can be various basic information when the vehicle is running, such as driving time information, travel information, speed information, turning information, longitude / latitude information, altitude information, and mobile phone call status information. , Sudden acceleration information, sudden deceleration information, sharp curve information, etc. can be included.

In the specific implementation process, the mobile SDK module in the sample driver's mobile terminal collects GPS, accelerometer and gyro sensor data in the sample driver's driving process, and acquires the sample driver's driving behavior data. , The driving behavior data of the sample driver can be uploaded to the server, and the driving behavior data of the sample driver and the identity of the sample driver can be bound and stored by the server.

In some embodiments, the preliminary interface can be used to obtain driving environment data for the sample driver. The driving environment data can be environmental information related to various driving activities in the environment in which the driving is located, and can include the link speed limit, the driving area, the road type, the terrain condition, the current weather condition of the driving, and the like. After acquiring the driving environment data of the sample driver by the preliminary interface, the driving environment data can be uploaded to the server, and the driving environment data of the sample driver and the identity of the sample driver can be bound and stored by the server.

In some embodiments, driving video data can be acquired by an on-board camera. The driving video data is a video of the driving period collected by the camera, and can include distance measurement information, lane information, road condition information, driving event, and the like. After the in-vehicle camera collects the driving video data, the driving video data can be uploaded to the server, and the server can bind and store the driving video data of the sample driver and the identity of the sample driver.

In some embodiments, the pre-existing interface allows sample driver sample insurance contract data to be obtained. The insurance contract data can be various information about the driver's insurance contract, and can include the driver's basic information, contract purchase information and the number of contract damage compensation cases, contract damage compensation amount, etc., but contract purchase The information includes information such as insurance coverage type and insurance coverage limit.

In the specific implementation process, the server can connect to the insurance contract system through a preliminary interface and obtain insurance contract data corresponding to the sample driver's identity from the business contract system based on the sample driver's identity. it can.

It should be explained that the above driver identity can be the driver's mobile phone number, username, ID card or other information that can uniquely label the driver.

In step S2, the current accumulation step of the sample data is determined based on the accumulated amount of the sample operation data and the sample insurance contract data.

In some embodiments, when the amount of sample insurance contract data accumulated is lower than the first threshold, the current accumulation staircase of sample data is determined as the first staircase. When the amount of sample insurance contract data accumulated is equal to or greater than the first threshold value, the current accumulation step of sample data is determined as the second step.

Current accumulation of sample data When the staircase is the first staircase, the amount of sample insurance contract data accumulated is low and insufficient to evaluate driving behavior. Therefore, it is possible to evaluate the current driving behavior of the driver using only the sample driving data. This type of method also allows the driver's driving behavior to be evaluated when there is not enough insurance policy data in the early stages of sample data collection.

When the current accumulation staircase of sample data is the second staircase, the accumulation amount of sample insurance contract data is relatively sufficient and can be used to evaluate driving behavior. Therefore, the current driver's driving behavior can be evaluated by using the sample driving data and the sample insurance contract data at the same time. With this type of method, it is possible to evaluate the driving behavior of the driver by adopting a more optimized driving evaluation method.

In some embodiments, the first staircase may include a third staircase in which the amount of sample insurance contract data accumulated is lower than the first threshold and the amount of sample operation data accumulated is lower than the second threshold. .. For example, the third staircase can be a cold start staircase where the sample data contains only a small amount of sample operation data and no sample insurance contract data. The so-called cold start staircase is a staircase that still activates the system operation when specific data is not acquired in the early stage of the system operation. In the present disclosure, the cold start staircase is a staircase that evaluates a driver's driving behavior based only on sample driving data when there is no sample insurance contract data.

In some embodiments, the first staircase may include a fourth staircase in which the amount of sample insurance contract data accumulated is lower than the first threshold and the amount of sample operation data accumulated is greater than or equal to the second threshold. it can. For example, the fourth staircase can be an initial staircase where the sample data contains sample driving data for multiple sample drivers and very little sample insurance contract data.

In some embodiments, the second step is a fifth step in which the amount of sample insurance contract data accumulated is greater than or equal to the first threshold and the amount of sample insurance contract data accumulated is lower than the amount of sample operation data accumulated. Can include stairs. For example, the fifth staircase can be a mid-term staircase in which the sample data includes sample driving data for a plurality of sample drivers and sample insurance contract data for some sample drivers.

In some embodiments, the second staircase is the sixth staircase in which the accumulated amount of sample insurance contract data is equal to or greater than the first threshold value and the accumulated amount of sample insurance contract data is equal to the accumulated amount of sample operation data. Can be included. For example, the sixth staircase can be a late staircase in which the sample data includes sample driving data of a plurality of sample drivers and corresponding sample insurance contract data.

Of course, the first threshold value and the second threshold value are predetermined values based on actual cases, and the present disclosure does not limit the threshold selection method.

Further, in the present disclosure, the accumulation amount of sample insurance contract data is lower than the accumulation amount of sample operation data, which includes sample operation data of a plurality of sample drivers in the sample data, but is a part of the plurality of sample drivers. It only contains sample insurance contract data for drivers. Further, the accumulation amount of the sample insurance contract data is equal to the accumulation amount of the sample operation data means that the sample data includes the sample operation data of a plurality of sample drivers and also includes the sample insurance contract data of the plurality of sample drivers. That is.

In the following description, the cold start staircase, the early staircase, the middle staircase, and the late staircase will be described by representing the third staircase, the fourth staircase, the fifth staircase, and the sixth staircase, respectively.

In step S3, the driving evaluation method corresponding to the currently accumulated stairs is used, and the accumulated sample data is used to evaluate the driving behavior of the current driver. However, driving behavior is evaluated using different driving evaluation methods on different stairs.

The driving behavior evaluation method provided by the embodiment of the present disclosure accumulates sample data, determines the currently accumulating staircase of the sample data, and uses the driving evaluation method corresponding to the currently accumulating staircase to evaluate the driving behavior of the current driver. evaluate. This makes it possible to provide a driving behavior assessment in the early stages of collecting driving data, and by gradually optimizing the parameters and model according to the accumulation of driving data and the addition of damage compensation data, the data accumulation differs. Driving risk can be estimated at any of the stairs.

In one embodiment of the present disclosure, when the staircase is currently a cold start staircase, the sample driver's driving data has just begun to accumulate, there is no sample driver's insurance contract data, and the rules and pre-weights Based on driving evaluation methods, the current driver's driving behavior can be evaluated. Specifically, as shown in FIG. 2, evaluating the driving behavior of the current driver using the driving evaluation method corresponding to the current accumulation staircase in step S3 may include the following steps. it can.

S311, the driving data of the current driver is acquired, and one type or a plurality of types of driving characteristics of the current driver are extracted.

However, the current driver's driving data can include at least one of driving behavior data, driving environment data and driving video data.

Specifically, the driving behavior data is acquired by the mobile SDK module, and the driving stroke, the driving time length, the maximum driving speed, the number of sudden accelerations, the number of sudden decelerations, and the number of sharp curves are obtained from the driving behavior data for each itinerary or unit time. It is possible to extract driving behavior characteristics including at least one of the number of sudden lane changes, whether or not the driver is tired, and whether or not the driver is driving in a dangerous time stage.

The driving environment data is acquired by the preliminary interface, and the number and length of speed overs for each itinerary or unit time, whether it is a dangerous environment, whether it is a familiar link, or whether it is bad weather from the driving environment data. It is possible to extract operating environment characteristics including at least one of them.

It is possible to acquire driving video data with an in-vehicle camera and extract driving video features from the driving video data, including at least one of the distance-speed ratio, whether it is in line or not, or whether it slows down when a forward event occurs. it can.

S312, distinguishing driving features into several dimensions, each dimension contains several types of driving features.

Specifically, the extracted plurality of types of driving features can be distinguished into several dimensions, such as speed, stroke, time, environment, driving behavior, etc., and each dimension includes some features. For example, the dimension "speed" can include several characteristics related to speed such as the number of speed overs, the length of overspeeds and the maximum driving speed, and the dimension "driving behavior" is the number of sudden accelerations. It can include several features related to driving operations such as the number of sudden decelerations, the number of sharp curves, the number of sudden lane changes, the distance / speed ratio, whether or not the vehicle is in line with the lane, and whether or not the vehicle decelerates when a forward event occurs.

Based on S313, the preliminary evaluation standard, the evaluation of each type of driving feature is acquired.

Specifically, an evaluation standard can be set for each type of driving feature, and the evaluation standard corresponding to each type of driving feature can be set differently based on a specific case.

For example, the evaluation standard of driving characteristics such as the number of overspeeds is
Score of the number of overspeeds = 100-Number of overspeeds * 10 / Driving stroke can be, the lower bound is 0, and on average 10 minutes is subtracted if there is one overspeed per kilometer. When, on average, the number of speed overs per kilometer> = 10, the driving feature score is 0.

S314, a weighting calculation is performed on the evaluation of each type of driving feature in each dimension to obtain the evaluation of each dimension.

Specifically, based on the feature weight value set for each type of driving feature, each term feature evaluation in each dimension can be weighted and summed to obtain an evaluation for each dimension, provided that the feature in each dimension is obtained. The sum of the weight values is 1.

The dimensional evaluation calculation formula is dimensional evaluation = Σ feature evaluation * feature weight value.

S315, a weighting calculation is performed on the evaluation of each dimension to obtain the current driver's driving behavior evaluation.

Specifically, based on the dimensional weight values set for each dimension, the weighting total can be performed for each dimensional evaluation to obtain the driving behavior evaluation of each driver, provided that the sum of the dimensional weight values is summed. Is 1.

The driving behavior evaluation calculation formula is driving behavior evaluation = Σ-dimensional evaluation * dimensional weight value.

In some embodiments, in steps S311 to S315 described with reference to FIG. 2 above, the current driver's driving characteristics are directly evaluated after obtaining an evaluation of the driving characteristics without making a dimension distinction. It may be used as an evaluation of the driving behavior of the current driver.

In the embodiment of the present disclosure, when the stairs are currently cold start stairs, the driving data of the sample driver has just begun to accumulate, there is no insurance contract data of the sample driver, and the driving evaluation method corresponding to the cold start stairs. By using to evaluate the driving behavior of the current driver, it is difficult to activate the supervision algorithm because the sample data at the initial stage of system operation is not sufficiently accumulated in the prior technology, and the driving behavior evaluation cannot be performed. Can solve the problem.

In one embodiment of the present disclosure, when the staircase is currently the initial staircase in the process of accumulating sample data, a certain amount of sample driver driving data is being accumulated and the sample driver's insurance contract data is still very high. The degree of relevance to the driver's driving data is relatively low as a whole, and the driving evaluation method based on the feature distribution function can be used to evaluate the driving behavior of the current driver. Specifically, as shown in FIG. 3, evaluating the driving behavior of the current driver using the driving evaluation method corresponding to the current accumulation staircase in step S3 may include the following steps. it can.

From S321, the sample driving data of each sample driver, a plurality of types of sample driving features of each sample driver are extracted.

Specifically, the process can be referred to in step S311 and is not repeated here.

S322, the distribution parameter of each type of sample driving feature is calculated based on one type or a plurality of types of sample driving feature of each sample driver.

Specifically, assuming that a distribution with sample driving characteristics is requested, the sample driving characteristics of all sample drivers are summarized, and the collected data is filtered including necessary outlier detection and data normalization. Based on the data fitting after filtering, the parameters of the requested distribution function of the sample driving feature can be obtained, and similarly, the distribution parameters of other sample driving features can be obtained.

が指数分布に依頼するとし、累積分布関数は、以下のようであり、但し、 For example, a sample driving feature, eg, the number of sudden accelerations per kilometer on average (outside 1).

Requests an exponential distribution, and the cumulative distribution function is as follows, but

。

..

全てのサンプル運転手の該運転特徴を纏め、必要な異常値検出、データの正規化を行い、フィルタリング後のデータフィッティングに基づいて該特徴を得、即ち平均的にキロメートル毎の急加速回数が依頼する分布関数のパラメータ
（外２）

を得る。

The driving characteristics of all sample drivers are summarized, necessary outlier detection, data normalization are performed, and the characteristics are obtained based on the data fitting after filtering, that is, the number of sudden accelerations per kilometer is requested on average. Parameters of the distribution function to be performed (outside 2)

To get.

S323, the driving data of the current driver is acquired, and one type or a plurality of types of driving characteristics of the current driver are extracted.

Specifically, the process can be referred to in step S311 and is not repeated here.

Based on the distribution parameters of each type of sample driving feature in S324, the cumulative distribution function of each type of driving feature of the current driver is calculated for each, and the evaluation corresponding to each type of driving feature is performed.

について、該運転特徴の依頼する指数分布のパラメータ
（外４）

に従って該運転特徴の累積分布関数値 For example, a driving feature that currently has a driver, such as the number of sudden accelerations per kilometer on average (outside 3).

The parameter of the exponential distribution requested by the driving feature (outside 4)

Cumulative distribution function value of the driving feature according to

を得て、それを該運転特徴に対応する評価とする。

Is obtained, and it is evaluated as corresponding to the driving feature.

S325, a weighting calculation is performed on the evaluation of each type of driving characteristic to obtain the current driver's driving behavior evaluation.

In the embodiments of the present disclosure, when the current staircase is the initial staircase, a certain amount of sample driver driving data is accumulated, and the sample driver's insurance contract data is still very small or scarce, and driving. The degree of relevance to the driving data of the hand is relatively low as a whole, and by evaluating the driving behavior of the current driver using the driving evaluation method corresponding to the initial stairs, a sample of the initial system operation in the prior technology Since the data accumulation is not sufficient, it is difficult to activate the supervision algorithm, and it is possible to solve the problem that the driving behavior evaluation cannot be performed.

In one embodiment of the present disclosure, when the staircase is currently a mid-term staircase in the process of accumulating sample data, a certain amount of sample driver driving data and some sample driver insurance contract data are accumulated. Therefore, it is possible to evaluate the driving behavior of the current driver by using the driving evaluation method based on the clustering analysis. Specifically, as shown in FIG. 4, the evaluation of the driving behavior of the current driver using the driving evaluation method corresponding to the current accumulation staircase in step S3 described above specifically involves the following steps. Can include.

Based on S331, sample insurance contract data, the driving evaluations of a plurality of sample drivers corresponding to the sample insurance contract data are acquired.

Specifically, the following operations can be performed on the insurance contract data of each sample driver.

Obtain the insurance claim value from the insurance contract data, determine the driving evaluation that has a mapping relationship with the insurance claim value based on the preliminary mapping relationship table, and use it as the sample driver's driving evaluation. In the mapping relationship table of, the higher the damage compensation rate, the lower the corresponding driving behavior evaluation.

However, the insurance claim value is the risk probability or compensation rate of the sample driver extracted from the insurance contract data of the sample driver, and the sample driving is performed from the mapping relationship table set in advance based on the risk probability or damage compensation rate. A hand driving behavior evaluation can be obtained and used as a sample driver's evaluation label. However, in the mapping relationship table, the higher the damage compensation rate, the lower the corresponding driving behavior evaluation. For example, the range of the evaluation score can be 0 to 100.

S332, the driving data of the plurality of sample drivers is clustered and analyzed, the centers of the plurality of clusters are acquired, each cluster is made into one risk level cluster, and each risk is based on the driving evaluation of the plurality of sample drivers. Calculate the operational rating of the level cluster.

However, clustering analysis can be performed on the driving data of some sample drivers using the K-means algorithm, and each cluster is regarded as one risk level cluster, the center of one cluster is acquired, and the same. It is possible that the drivers belonging to the same category have the same or similar risk level, and the average value of the driving evaluation of the sample driver in the same category is used as the driving evaluation of the category (immediately, the risk level cluster). it can.

S333, the driving data of the current driver is acquired, and the similarity distance between the driving data of the current driver and the centers of a plurality of clusters is calculated.

Specifically, the embodiments of the present disclosure do not limit the specific calculation process.

Based on the calculation result of the similarity distance in S334, the risk level cluster corresponding to the current driver is determined, and the driving evaluation of the determined risk level cluster is set as the driving evaluation of the current driver.

Specifically, the risk level cluster corresponding to the minimum value in the calculation result of the similarity distance is defined as the risk level cluster corresponding to the current driver.

In the embodiment of the present disclosure, when the staircase is currently a medium-term staircase in the process of accumulating sample data, a certain amount of sample driver's driving data and some sample driver's insurance contract data are accumulated. By using the driving evaluation method corresponding to the medium-term stairs, it is possible to evaluate the driving behavior of the driver at present, and it is difficult to start the supervision algorithm because the sample data at the initial stage of system operation is not sufficiently accumulated in the preceding technology. It is possible to solve the problem that the driving behavior evaluation cannot be performed.

In one embodiment of the present disclosure, when the current staircase is a late staircase, the sample driver's driving data and damage compensation data are relatively aligned, and the driver's driving data and damage compensation data are relatively related. It is possible to evaluate the driving behavior of the current driver using the driving evaluation method based on the evaluation model after training. Specifically, as shown in FIG. 5, the evaluation of the driving behavior of the current driver by using the driving evaluation method corresponding to the current accumulation staircase in the above-mentioned step S3 specifically involves the following steps. Can include.

S341, Based on the sample driver's insurance contract data, the driving evaluations of a plurality of sample drivers corresponding to the sample insurance contract data are acquired, and the sample driving features of the plurality of sample drivers are extracted from the sample driving data. ..

Specifically, in this step, the process of acquiring the driving evaluations of a plurality of sample drivers corresponding to the sample insurance contract data based on the sample driver's insurance contract data is the same as in step S331, and here. Do not repeat.

In this step, the process of extracting the sample driving features of the plurality of sample drivers from the sample driving data is the same as in step S311 and is not repeated here.

Based on S342, the driving characteristics of each sample driver, and the driving evaluation of each sample driver, a driving behavior evaluation model is built.

Specifically, for each sample driver's driving characteristics and each sample driver's driving evaluation, a driving behavior evaluation model is built by traditional machine learning or deep learning method, and the driving behavior evaluation model is used as an online driving behavior evaluation. Can be stored offline so that it can be recalled at.

However, a driving behavior evaluation model can be built using methods such as linear regression, random forest, decision tree, xgboss, etc., and the present disclosure does not limit this.

S343, the driving behavior evaluation model is used to evaluate the current driving behavior of the driver.

Specifically, the driving characteristics of the current driver are extracted from the driving data of the current driver, the driving characteristics of the current driver are input to the driving behavior evaluation model, and the driving behavior evaluation of the current driver is obtained and output.

In the embodiment of the present disclosure, when the current staircase is a late staircase in the process of accumulating sample data, the sample driver's driving data and damage compensation data are relatively arranged, and the driver's driving data and damage compensation data are combined. Since the relationships are relatively complete, it is possible to obtain a driving evaluation model by training by adopting device learning or deep learning, and further evaluate the driving behavior of the current driver.

FIG. 6 shows a typical layout block diagram of the driving behavior evaluation device according to another embodiment of the present disclosure. The driving behavior evaluation device provided by the embodiment of the present disclosure can be used to execute the driving behavior evaluation method in the above embodiment.

In some embodiments, the device 600 may include a processing circuit 60. The processing circuit 60 of the device 600 provides various functions of the device 600. In some embodiments, the processing circuit 60 of the apparatus 600 can be arranged to perform the driving behavior evaluation method described above with reference to FIG.

The processing circuit 60 can be various realizations of a digital circuit system, an analog circuit system, or a mixed signal (combination of analog and digital) circuit systems that perform functions in a computing system. Processing circuits include, for example, integrated circuits (ICs), dedicated integrated circuits (ASICs), parts or circuits of a single processor core, whole processor cores, a single processor, such as a field programmable gate array (FPGA). It can include programmable hardware equipment and / or a system that includes multiple processors.

In some embodiments, the processing circuit 60 can include a storage module 61, a confirmation module 62, and an evaluation module 63.

The storage module 61 is arranged so as to store sample data including sample driving data and sample insurance contract data of the sample driver, and the confirmation module 62 is based on the storage amount of the sample driving data and the sample insurance contract data. Therefore, the evaluation module 63 is arranged so as to determine the current storage staircase of the sample data, and the evaluation module 63 uses the operation evaluation method corresponding to the current storage staircase and uses the accumulated sample data to be used by the current driver. It is arranged to evaluate the driving behavior of. The modules 61 to 63 can be arranged so as to execute steps S1 to S3 in the driving behavior evaluation method shown in FIG. 1 above.

In some embodiments, the device 600 may include memory (not shown). The memory of the device 600 can store the information generated by the processing circuit 60 and the programs and data used for the operation of the device 600. The memory can be a volatile memory and / or a non-volatile memory. For example, memory can 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. Further, the device 600 can be realized at the chip level, or can be realized at the equipment level including other peripheral members.

It should be understood that each of the above modules is only a logic module that is distinguished based on the specific function realized by it, and is not intended to limit the specific implementation method. When actually realized, each of the above modules can be realized as an independent physical entity, or may be realized by a single entity (for example, a processor (CPU or DSP, etc.), an integrated circuit, etc.). ..

The driving behavior evaluation device provided by the embodiment of the present disclosure belongs to the same concept of the invention as the driving behavior evaluation method provided by the embodiment of the present disclosure, and the driving behavior provided by any of the embodiments of the present disclosure. The evaluation method can be performed and has a functional module and a beneficial effect corresponding to the method of performing the driving behavior evaluation. For technical details not described in detail in this embodiment, the driving behavior evaluation method provided by the embodiments of the present disclosure can be referred to and is not repeated here.

FIG. 7 shows a typical arrangement of computing equipment 700 that can realize the embodiments of the present disclosure. The calculation equipment 700 is an example of hardware equipment to which the above technical proposal of the present disclosure can be applied. Computational equipment 700 can be any instrument arranged to perform processing and / or computation. The computing facility 700 can be a workstation, a server, a desktop computer, a notebook computer, a tablet computer, a personal data assistant (PDA), a smartphone, an in-vehicle computer, or a combination of the above, but is not limited to this, as shown in FIG. In addition, the computing facility 700 may include one or more components that connect or communicate with the bus 702 via one or more interfaces. Bus 702 is an industry standard architecture (ISA) bus, a Micro Channel Architecture (MCA) bus, an enhanced ISA (EISA) bus, a video electronics standardization association (VESA) local bus, and a peripheral device. Connection (PCI) buses and the like can be included, but are not limited to this. The computing equipment 700 can include, for example, one or more processors 704, one or more input equipment 706, and one or more output equipment 708. One or more processors 704 can be any type of processor and can include, but are not limited to, one or more general or dedicated processors (eg, dedicated processing chips). The processor 704 can correspond to, for example, the processing circuit 60 in FIG. 6, and is arranged so as to realize the function of each module of the apparatus for inspecting the certificate and the bearer of the present disclosure. The input equipment 706 can be any type of input equipment capable of inputting information into the computing equipment and can include a mouse, keyboard, touch screen, microphone and / or remote controller. Not limited to. Output equipment 708 can be any type of equipment capable of representing information and can include, but is not limited to, displays, speakers, video / audio output terminals, vibrators and / or printers. ..

Computer equipment 700 may include or be connected to non-volatile storage 714, which non-volatile storage 714 can be any non-volatile and data storage capable storage and disk drive. , Optical storage, solid state memory, floppy disk, flexible disk, hard disk, magnetic tape or any other magnetic medium, compressed disk or any other optical medium, cache memory and / or any other memory chip Or the module and / or any other medium from which the computer can read data, commands and / or codes can be included, but not limited to. The computing facility 700 may include a random access memory (RAM) 710 and a read-only memory (ROM) 712. The ROM 712 can store a program, a practical program or a course executed in a non-volatile manner. The RAM 710 provides storage of volatile data and can store commands relating to the operation of the computing facility 700. The computing facility 700 may include a network / bus interface 716 connected to the data link 718. The network / bus interface 716 can be any type of equipment or system capable of communicating with peripherals and / or networks, and can be a modem, network card, infrared communication equipment, wireless communication equipment and / or chip. It can include, but is not limited to, a set (eg, Bluetooth® equipment, 802.11 equipment, WiFi equipment, WiMax equipment, cellular communication equipment, etc.).

It should be noted that another embodiment of the present disclosure is a driving behavior evaluation device.
With one or more processors
Includes a storage device for storing one or more programs,
When the one or more programs are executed by the one or more processors, the driving behavior evaluation device that makes the one or more processors realize the driving behavior evaluation method described in the above-described embodiment. Further provide.

In addition, another embodiment of the present disclosure is a storage medium that can be read by a computer that stores a computer program, and when the program is executed by a processor, the operation described in the above-described embodiment is performed. Further providing a computer-readable storage medium that realizes the behavior evaluation method.

It should be explained that in the description of the present disclosure, the terms "first", "second", etc. are for illustration purposes only and do not indicate or imply relative importance and order.

For those skilled in the art, the embodiments in the present disclosure can provide methods, systems, or computer program products. Therefore, in the examples of the present disclosure, the form of an example of all hardware, an example of all software, or an example of a union of software and hardware can be adopted. Further, in the embodiments of the present disclosure, a computer-usable storage medium (including, but not limited to, disk memory, CD-ROM, optical memory, etc.) containing a program code that can be used by one or more computers. The format of the computer program product implemented above can be adopted.

The embodiments of the present disclosure are described with reference to the methods, equipment (systems), and flow diagrams and / or block diagrams of the computer program products of the embodiments of the present disclosure. It should be understood that computer program directives can implement flow and / or block-by-block and flow and / or block unions in flow diagrams and / or block diagrams. These computer program instructions are provided to a general computer, a dedicated computer, an embedded processor or the processor of another programmable data processing facility to generate one device, which is executed by the processor of the computer or other programmable data processing facility. The command may form a device for realizing the function specified in one flow or a plurality of flows and / or a block in the block diagram.

By storing these computer program commands in a computer-readable memory capable of operating the computer or other programmable data processing equipment in a specific manner, the instructions stored in the computer-readable memory can be obtained. , One flow or multiple flows in the flow diagram and / or one block or plurality of blocks in the block diagram may form a product containing a command device that realizes the specified function.

By loading those computer program instructions into a computer or other programmable data processing facility, the computer or other programmable data processing facility can perform a series of operational steps on the computer or other programmable facility to generate the processing that the computer implements. Instructions executed on other programmable equipment provide steps to realize the functions specified in one flow or multiple flows and / or block diagrams in a flow diagram.

Although preferred embodiments of the embodiments of the present disclosure have been described, other modifications and modifications can be made to those embodiments once those skilled in the art have learned the basic concepts of inventive step. Accordingly, the appended claims are intended to be construed as including all modifications and modifications contained within the preferred embodiments and embodiments of the present disclosure.

Obviously, one of ordinary skill in the art can make various changes and modifications to the present disclosure without departing from the spirit and scope of the present disclosure. Thus, where these modifications and modifications of the invention are within the claims of the present disclosure and the equivalent technology thereof, the present disclosure is also intended to include those modifications and modifications.

Claims (18)

It is a driving behavior evaluation method
Accumulating sample data including sample driver data and sample insurance contract data,
To determine the current accumulation staircase of the sample data based on the accumulated amount of the sample operation data and the sample insurance contract data.
A driving behavior evaluation method comprising using the driving evaluation method corresponding to the currently accumulated stairs and evaluating the driving behavior of the current driver using the accumulated sample data. When the accumulated amount of the sample insurance contract data is lower than the first threshold value, the currently accumulated staircase of the sample data is determined as the first staircase.
The first aspect of claim 1, wherein the driving evaluation method corresponding to the first staircase is used, and the driving behavior of the current driver is evaluated by using the sample driving data in the accumulated sample data. Method. When the accumulated amount of the sample insurance contract data is equal to or greater than the first threshold value, the currently accumulated staircase of the sample data is determined as the second staircase.
A claim characterized in that the driving evaluation method corresponding to the second staircase is used, and the driving behavior of the current driver is evaluated by using the sample driving data and the sample insurance contract data in the accumulated sample data. The method described in 1. The first staircase includes a third staircase in which the accumulated amount of the sample insurance contract data is lower than the first threshold value and the accumulated amount of the sample operation data is lower than the second threshold value.
When the present accumulation staircase is the third staircase, assessing the driving behavior of the current driver
To acquire the driving data of the current driver and extract the driving characteristics of the current driver.
The method according to claim 2, wherein the evaluation of the driving characteristics of the current driver is acquired as the driving behavior evaluation of the current driver based on a preliminary evaluation standard. The first staircase includes a fourth staircase in which the accumulated amount of the sample insurance contract data is lower than the first threshold value and the accumulated amount of the sample operation data is equal to or more than the second threshold value.
When the present accumulation staircase is the fourth staircase, assessing the driving behavior of the current driver
Extracting the sample driving characteristics of the sample driver from the sample driving data of the sample driver,
To calculate the distribution parameters of the sample driving feature based on the sample driving feature of the sample driver.
To acquire the driving data of the current driver and extract the driving characteristics of the current driver.
The second aspect of claim 2, wherein the cumulative distribution function of the driving characteristics of the current driver is calculated as an evaluation of the driving behavior of the current driver based on the distribution parameters of the sample driving characteristics. Method. The second step includes a fifth step in which the accumulated amount of the sample insurance contract data is equal to or more than the first threshold value and the accumulated amount of the sample insurance contract data is lower than the accumulated amount of the sample operation data. ,
When the present accumulation staircase is the fifth staircase, assessing the driving behavior of the current driver
Obtaining driving evaluations of a plurality of sample drivers corresponding to the sample insurance contract data based on the sample insurance contract data.
The driving data of the plurality of sample drivers is clustered and analyzed, the centers of the plurality of clusters are acquired, each cluster is made into one risk level cluster, and each risk level cluster is based on the driving evaluation of the plurality of sample drivers. To calculate the driving evaluation of
To acquire the driving data of the current driver and calculate the similarity distance between the driving data of the current driver and the centers of the plurality of clusters.
Including that the risk level cluster corresponding to the current driver is determined based on the calculation result of the similarity distance, and the driving evaluation of the determined risk level cluster is used as the driving evaluation of the current driver. The method according to claim 3. The second step includes a sixth step in which the accumulated amount of the sample insurance contract data is equal to or more than the first threshold value and the accumulated amount of the sample insurance contract data is equal to the accumulated amount of the sample operation data. ,
When the present accumulation staircase is the sixth staircase, assessing the driving behavior of the current driver
Obtaining driving evaluations of a plurality of sample drivers corresponding to the sample insurance contract data based on the sample insurance contract data.
Extracting sample driving features of the plurality of sample drivers from the sample driving data,
Building a driving behavior evaluation model based on the sample driving characteristics of the plurality of sample drivers and the corresponding driving evaluations.
The method according to claim 3, wherein the driving behavior evaluation model is used to evaluate the current driver's driving behavior. Obtaining a driving evaluation of a plurality of sample drivers corresponding to the contract data based on the sample insurance contract data described above may be performed.
For sample insurance contract data of each of the plurality of sample drivers
Obtaining insurance claim figures from the sample insurance contract data,
According to claim 6 or 7, the driving evaluation having a mapping relationship with the insurance claim value is determined as the driving evaluation of the sample driver based on the preliminary mapping relationship table. The method described. It is a driving behavior evaluation device
A storage module arranged to store sample data, including sample driver data and sample insurance contract data, and a storage module.
Based on the accumulated amount of the sample operation data and the sample insurance contract data, the confirmation module arranged so as to determine the current accumulation step of the sample data, and
The driving behavior is characterized by including an evaluation module arranged to evaluate the current driver's driving behavior using the accumulated sample data using the driving evaluation method corresponding to the currently accumulated staircase. Evaluation device. When the accumulated amount of the sample insurance contract data is lower than the first threshold value, the currently accumulated staircase of the sample data is determined as the first staircase.
The ninth aspect of claim 9, wherein the driving evaluation method corresponding to the first staircase is used, and the driving behavior of the current driver is evaluated by using the sample driving data in the accumulated sample data. apparatus. When the accumulated amount of the sample insurance contract data is equal to or greater than the first threshold value, the currently accumulated staircase of the sample data is determined as the second staircase.
A claim characterized in that the driving evaluation method corresponding to the second staircase is used, and the driving behavior of the current driver is evaluated by using the sample driving data and the sample insurance contract data in the accumulated sample data. 9. The apparatus according to 9. The first staircase includes a third staircase in which the accumulated amount of the sample insurance contract data is lower than the first threshold value and the accumulated amount of the sample operation data is lower than the second threshold value.
When the current storage staircase is the third staircase, the evaluation module
The driving data of the current driver is acquired, and the driving characteristics of the current driver are extracted.
The device according to claim 10, wherein the device is arranged so as to acquire the evaluation of the driving characteristics of the current driver as the driving behavior evaluation of the current driver based on a preliminary evaluation standard. The first staircase includes a fourth staircase in which the accumulated amount of the sample insurance contract data is lower than the first threshold value and the accumulated amount of the sample operation data is equal to or more than the second threshold value.
When the current storage staircase is the fourth staircase, the evaluation module
From the sample driving data of the sample driver, the sample driving characteristics of the sample driver are extracted.
Based on the sample driving characteristics of the sample driver, the distribution parameters of the sample driving characteristics are calculated.
The driving data of the current driver is acquired, and the driving characteristics of the current driver are extracted.
The tenth aspect of claim 10, wherein the cumulative distribution function of the driving characteristics of the current driver is arranged so as to be calculated as a driving behavior evaluation of the current driver based on the distribution parameters of the sample driving characteristics. The device that was used. The second step includes a fifth step in which the accumulated amount of the sample insurance contract data is equal to or more than the first threshold value and the accumulated amount of the sample insurance contract data is lower than the accumulated amount of the sample operation data. ,
When the current storage staircase is the fifth staircase, the evaluation module
Based on the sample insurance contract data, the driving evaluations of a plurality of sample drivers corresponding to the sample insurance contract data are acquired.
The driving data of the plurality of sample drivers is clustered and analyzed, the centers of the plurality of clusters are acquired, each cluster is made into one risk level cluster, and each risk level cluster is based on the driving evaluation of the plurality of sample drivers. Calculate the driving evaluation of
The driving data of the current driver is acquired, and the similarity distance between the driving data of the current driver and the centers of the plurality of clusters is calculated.
Based on the calculation result of the similarity distance, the risk level cluster corresponding to the current driver is determined, and the determined driving evaluation of the risk level cluster is arranged so as to be the driving evaluation of the current driver. The device according to claim 11, wherein the device is provided. The second step includes a sixth step in which the accumulated amount of the sample insurance contract data is equal to or more than the first threshold value and the accumulated amount of the sample insurance contract data is equal to the accumulated amount of the sample operation data. ,
When the current storage staircase is the sixth staircase, the evaluation module
Based on the sample insurance contract data, the driving evaluations of a plurality of sample drivers corresponding to the sample insurance contract data are acquired.
From the sample driving data, sample driving features of the plurality of sample drivers are extracted.
Based on the sample driving characteristics of the plurality of sample drivers and the corresponding driving evaluations, a driving behavior evaluation model was constructed.
The device according to claim 11, wherein the device is arranged to evaluate the current driver's driving behavior using the driving behavior evaluation model. Obtaining a driving evaluation of a plurality of sample drivers corresponding to the contract data based on the sample insurance contract data described above may be performed.
For sample insurance contract data of each of the plurality of sample drivers
Obtaining insurance claim figures from the sample insurance contract data,
According to claim 14 or 15, the driving evaluation having a mapping relationship with the insurance claim value is determined as the driving evaluation of the sample driver based on the preliminary mapping relationship table. The device described. It is a driving behavior evaluation device
With one or more processors
Includes a storage device for storing one or more programs,
When the one or more programs are executed by the one or more processors, the one or more processors realizes the driving behavior evaluation method according to any one of claims 1 to 8. A characteristic driving behavior evaluation device. A storage medium that can be read by a computer that stores computer programs.
A computer-readable storage medium comprising implementing the driving behavior evaluation method according to any one of claims 1 to 8, when the program is executed by a processor.

Images