Driving behavior data processing method and apparatus, and device

Abstract

A driving behavior data processing method and apparatus, and a device. The method comprises: on the basis of a travelled distance of a vehicle, acquiring a trip segment to be measured; on the basis of a preset driving dimension, performing screening on driving data of said trip segment to obtain target driving data corresponding to the preset driving dimension, wherein the preset driving dimension includes a dimension of a driving environment category and / or a dimension of a driving performance category; on the basis of a preset dimension evaluation model, processing the target driving data to obtain a predicted score, wherein the predicted score is used for representing a driving behavior score of feature data in the target driving data in the preset driving dimension; and on the basis of the position of said trip segment in the travelled distance and the predicted score, acquiring a target score of said trip segment, wherein the target score is used for representing a driving behavior score of the vehicle in the preset driving dimension within said trip segment. The method improves the accuracy of driving behavior analysis.

Description

Driving behavior data processing method, device and equipment

[0001] This disclosure claims priority to Chinese patent application number 202410189773.2, filed with the Patent Office of China on February 20, 2024, entitled “Driving Behavior Data Processing Method, Device and Equipment,” the entire contents of which are incorporated herein by reference. Technical Field

[0002] The present disclosure relates to the technical field of driving data processing, and in particular to a method, device and apparatus for processing driving behavior data. Background Art

[0003] The evaluation of driving behavior helps to improve driving safety and reduce driving energy consumption, and the evaluation results can also serve as a reference for vehicle insurance assessment.

[0004] Current driving behavior assessment schemes generally directly obtain all data from the driver's use of the vehicle, then analyze the data based on safe driving specifications or energy-saving driving specifications, and deduct points for behaviors in the data that do not comply with safe driving specifications or energy-saving driving specifications, thereby obtaining a driving behavior assessment score. However, the assessment data used in this method is data from the driving process, and does not take into account the driving environment during the driving process. The driving environment will have an impact on the driving data. Therefore, the obtained assessment score is not accurate and cannot truly and accurately reflect the driver's driving style. Summary of the Invention

[0005] The present disclosure provides a driving behavior data processing method, device and apparatus to solve the problem of low accuracy of driving behavior analysis results.

[0006] In a first aspect, the present disclosure provides a driving behavior data processing method, comprising: obtaining a to-be-tested trip segment based on a vehicle's mileage; screening target driving data corresponding to a preset driving dimension in the driving data of the to-be-tested trip segment based on a preset driving dimension; wherein the preset driving dimension includes a dimension of a driving environment category and / or a dimension of a driving performance category; processing the target driving data based on a preset dimension evaluation model to obtain a predicted score; wherein the predicted score is used to characterize a driving behavior score of feature data in the target driving data in the preset driving dimension; obtaining a target score for the to-be-tested trip segment based on a position of the to-be-tested trip segment in the mileage and the predicted score, wherein the target score is used to characterize the driving behavior score of the vehicle in the preset driving dimension in the to-be-tested trip segment.

[0007] In one implementable manner, the preset driving dimension includes a dimension of a driving performance category, and the dimension of the driving performance category includes a stability dimension; based on the preset driving dimension, screening the target driving data corresponding to the preset driving dimension from the driving data of the travel segment to be tested includes: obtaining the acceleration data of the vehicle in the travel segment to be tested; obtaining the number of acceleration behaviors in the acceleration data whose value is greater than a preset acceleration value, and using the number of acceleration behaviors as the target driving data on the stability dimension.

[0008] In one implementable manner, the preset driving dimension includes a dimension of a driving performance category, the dimension of the driving performance category includes a stability dimension, and the target driving data includes the number of acceleration behaviors in which the value of the acceleration data of the vehicle is greater than a preset acceleration value; the target driving data is processed based on the preset dimensional evaluation model to obtain a prediction score, including: based on the acceleration direction corresponding to the acceleration behavior number, screening the acceleration behavior sub-numbers of different acceleration directions in the acceleration behavior number; according to the preset dimensional evaluation models for different acceleration directions, the acceleration behavior sub-numbers of the corresponding acceleration directions are processed to obtain a prediction score for the corresponding acceleration direction.

[0009] In one possible implementation, the dimension evaluation model preset for different acceleration directions is: μ x =a x +b x *n x

[0010] Where μ is the prediction score, a is the first hyperparameter, b is the second hyperparameter, n is the number of acceleration behaviors, different acceleration directions correspond to the first and second hyperparameters, and x is the sequence number of the acceleration direction.

[0011] In one implementation, obtaining a target score for the trip segment to be measured based on the position of the trip segment to be measured in the mileage and the predicted score includes: obtaining an initial trip segment; wherein the initial trip segment is located before the trip segment to be measured in the mileage of the vehicle; if the initial trip segment does not exist, obtaining a first score based on the predicted score, and using the first score as the target score; if the initial trip segment exists, obtaining a second score based on the predicted score, and obtaining the target score based on the difference between a third score of the initial trip segment and the second score, wherein the ratio between the first score and the second score is a preset weighted value.

[0012] In one implementable manner, the preset driving dimension includes a dimension of a driving performance category, and the dimension of the driving performance category includes a stability dimension; obtaining the first score based on the predicted score includes: obtaining the predicted score and target driving data; wherein the target driving data includes acceleration behavior sub-numbers in different acceleration directions, and the predicted score includes the predicted scores of acceleration behavior sub-numbers in different acceleration directions, and the acceleration direction includes longitudinal and lateral; obtaining the first score based on the predicted score and the target driving data based on a preset target evaluation model; wherein the target evaluation model is: Score s =Max(0,(L-n1*μ1-n2*μ2))*δ

[0013] Among them, Score s is the first score, n1 is the number of acceleration behaviors in the longitudinal positive acceleration direction, n2 is the number of acceleration behaviors in the longitudinal negative acceleration direction, n3 is the number of acceleration behaviors in the lateral acceleration direction, δ is the preset weighted value, μ1 is the predicted score of the longitudinal acceleration behavior sub-number, μ2 is the predicted score of the lateral acceleration behavior sub-number, and L is the preset scoring threshold.

[0014] In one implementation, obtaining the target score based on the difference between the third score and the second score of the initial trip segment includes: if the difference between the third score and the second score is within a preset numerical range, using the second score evaluation score as the target driving behavior evaluation score; if the difference between the third score and the second score is not within the preset numerical range, obtaining the target score based on the third score and a preset error value.

[0015] In a second aspect, the present disclosure provides a driving behavior data processing device, comprising: a module for determining a travel segment to be measured, configured to obtain the travel segment to be measured based on the mileage of a vehicle; a target driving data acquisition module, configured to filter target driving data corresponding to a preset driving dimension in the driving data of the travel segment to be measured based on a preset driving dimension; wherein the preset driving dimension includes a dimension of a driving environment category and / or a dimension of a driving performance category; a predicted score acquisition module, configured to process the target driving data based on a preset dimension evaluation model to obtain a predicted score; wherein the predicted score is used to characterize the driving behavior score of the feature data in the target driving data in the preset driving dimension; a target score acquisition module, configured to obtain a target score for the travel segment to be measured based on the position of the travel segment to be measured in the mileage and the predicted score, wherein the target score is used to characterize the driving behavior of the vehicle in the preset driving dimension in the travel segment to be measured.

[0016] In a third aspect, the present disclosure provides an electronic device, comprising: a processor and a memory; the memory is used to store instructions; the processor is used to execute the instructions in the memory, so that the electronic device performs the driving behavior data processing method described in the first aspect.

[0017] In a fourth aspect, the present disclosure provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement the driving behavior data processing method as described in the first aspect.

[0018] In a fifth aspect, the present disclosure provides a computer program product, comprising a computer program, which, when executed by a processor, implements the driving behavior data processing method as described in the first aspect.

[0019] In a sixth aspect, the present disclosure provides a vehicle comprising the driving behavior data processing device as described in the second aspect.

[0020] The driving behavior data processing method, device, and apparatus provided by the present disclosure screen target driving data of corresponding dimensions based on the dimensions of the driving environment category and the dimensions of the driving performance category, and use this target driving data as data that accurately characterizes the corresponding dimensions of the vehicle in the travel segment to be tested, thereby reducing the problem of excessive data noise during driving behavior data processing. At the same time, feature data in the target driving data is processed using a dimensional evaluation model to obtain a predicted score, thereby efficiently and accurately obtaining target scores representing driving behavior scores in different driving dimensions based on the predicted score and the position of the travel segment to be tested. This process requires no human intervention and is low-cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0022] FIG1 is a diagram illustrating an implementation scenario of a method for processing driving behavior data according to an embodiment;

[0023] FIG2 is a flow chart of a method for processing driving behavior data according to an embodiment;

[0024] FIG3 is a flowchart of a method for processing driving behavior data according to another embodiment;

[0025] FIG4 is a schematic diagram of a parallel driving scenario according to an embodiment;

[0026] FIG5 is a schematic diagram of a close-following vehicle driving scenario according to an embodiment;

[0027] FIG6 is a schematic diagram of an overtaking scenario shown in an embodiment;

[0028] FIG7 is a schematic diagram of a successful overtaking scenario according to an embodiment;

[0029] FIG8 is a schematic diagram of an overtaking failure scenario shown in an embodiment;

[0030] FIG9 is a schematic diagram of an aggressive lane change scenario according to an embodiment;

[0031] FIG10 is a flow chart of a method for processing driving behavior data according to another embodiment;

[0032] FIG11 is a flow chart of a method for processing driving behavior data according to another embodiment;

[0033] FIG12 is a structural diagram of a driving behavior data processing device according to an embodiment;

[0034] FIG13 is a block diagram of an electronic device according to an embodiment.

[0035] The above drawings illustrate specific embodiments of the present disclosure, which will be described in more detail below. These drawings and textual descriptions are not intended to limit the scope of the present disclosure in any way, but rather to illustrate the concepts of the present disclosure to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0036] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.

[0037] First, the terms involved in this disclosure are explained:

[0038] HMI: Human Machine Interface.

[0039] Tcam: Telematics & Connectivity Antenna Module, smart antenna module for Internet of Vehicles.

[0040] ADMC: AD Main Computer Unit, autonomous driving central processor.

[0041] DHU: Display Head Unit, driving information and entertainment host.

[0042] g-value: A customary unit used to describe vehicle acceleration, 1g = 9.8m / s^2 (square meters per second).

[0043] MQTT: A client-server based publish / subscribe messaging transport protocol.

[0044] ADAS: Advanced Driver Assistance System, driving assistance system.

[0045] Currently, driving behavior scores are mostly assessed from the perspectives of safe driving and energy conservation and environmental protection through basic data such as vehicle speed. Most of them use a deduction system to count and deduct points for violations of safe or energy-saving driving regulations to obtain the final driving evaluation.

[0046] For example, a driving behavior score assessment system may improve the score by reducing the number of sudden accelerations during driving; reduce the number of sudden decelerations during driving; and consider fatigue driving if continuous driving for more than four hours is considered fatigue driving, and the longer the fatigue driving time, the more points will be reduced; and consider the proportion of fast charging during driving if appropriate use of slow charging will improve the score.

[0047] It can be seen that the evaluation of this driving behavior score is very concrete. It is aimed at a certain car use behavior alone, and the evaluation mechanism is simple and single. It is more like a driving behavior standard to guide the driver, and cannot truly reflect the driver's driving style.

[0048] The driving behavior data processing method provided by the present disclosure can more comprehensively reflect the driver's driving behavior preferences and driving style with high accuracy.

[0049] The following detailed description of the technical solution of the present disclosure and how the technical solution of the present disclosure solves the above-mentioned technical problems is provided with specific embodiments. The following specific embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. The following embodiments of the present disclosure are described in conjunction with the accompanying drawings.

[0050] In the technical solution of this application, the collection, storage, use, processing, transmission, provision and disclosure of driving behavior data, driving data and other information involved comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0051] It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned. They should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0052] The driving behavior data processing method of this embodiment can be applied to a driving behavior data processing device, which is built into a vehicle-mounted system or disposed in other systems that are communicatively connected to the vehicle-mounted system, such as a cloud system.

[0053] It can be understood that the unit of speed in this embodiment is km / h (kilometers per hour), the unit of time is s (seconds), and the unit of acceleration is g (1g=9.8m / s^2).

[0054] As shown in Figure 1, the implementation scenario diagram of the driving behavior data processing method includes an ADMC module 1, a Tcam module 2, a DHU module 3, a data warehouse 4, and a driving behavior data processing device 5. The ADMC module 1, the Tcam module 2, and the DHU module 3 can be set on the vehicle side, while the data warehouse 4 and the driving behavior data processing device 5 can be set on the cloud.

[0055] Among them, the ADMC module 1 is set in the vehicle to be processed for driving behavior data, and is used to obtain relevant data of the vehicle during driving, such as the vehicle's position data, speed data, etc., for subsequent driving behavior data processing. The Tcam module 2 obtains data from other vehicles. If the vehicle to be processed for driving behavior data is taken as the first vehicle, the second data acquisition module 2 obtains data of other vehicles other than the first vehicle during the driving process of the first vehicle, such as the position relationship between other vehicles and the first vehicle, the speed of other vehicles and other data.

[0056] The data acquired by the ADMC module 1 can be transmitted to the DHU module 3 via the FR bus (a transmission mode with frames as data units) or Ethernet. The data in the DHU module 3 enters the data warehouse 4 through MQTT. At the same time, the data in the Tcam module 2 also enters the data warehouse 4, so that the driving behavior data processing device 5 performs driving behavior data processing based on the data in the data warehouse 4. For example, the data warehouse 4 obtains relevant data in a certain travel segment to be tested, and then the driving behavior data processing device 5 performs driving behavior data processing based on the data in the data warehouse 4 to obtain the vehicle's driving behavior score in the travel segment to be tested. The score can be displayed through the vehicle's HMI (Human Machine Interface) interface.

[0057] FIG2 is a flow chart of a method for processing driving behavior data according to an embodiment. Referring to FIG2 , the method for processing driving behavior data may include steps S210 to S270, which are described in detail as follows:

[0058] Step S210: obtaining a travel segment to be measured based on the vehicle's mileage.

[0059] In this embodiment, the travel segment to be measured is a certain section of the vehicle's entire travel mileage, such as a certain travel segment of the most recent mileage (such as 100 kilometers or 150 kilometers), or a certain travel segment of the most recent duration (such as a week or a month). Of course, it can also be a certain travel segment of a certain mileage in a certain period or a certain duration. In some embodiments, the travel segment to be measured can be a travel segment to be measured consisting of a certain type of driving environment according to different preset dimensions. For example, for a curve, the travel segment to be measured can also be a travel segment including a certain number of curve sections, such as taking the travel of the most recent 100 curve sections as the travel segment to be measured. No specific restriction is made here, and the travel segment to be measured can be limited to a travel segment with a travel length and a travel duration.

[0060] Step S230: Based on the preset driving dimension, target driving data corresponding to the preset driving dimension is filtered from the driving data of the travel segment to be measured.

[0061] Among them, the preset driving dimensions include dimensions of driving environment categories and / or dimensions of driving performance categories; the driving environment categories may include curve dimensions, and the target driving data of the curve dimensions are obtained accordingly; the driving performance categories may include speed dimensions, movement dimensions, stability dimensions, braking dimensions, and defensive dimensions, etc. In this embodiment, target driving data corresponding to one or more preset driving dimensions can be obtained, and each preset driving dimension corresponds to one target driving data. Then, the driving behavior data are processed separately for the target driving data of each preset driving dimension to obtain the target score of the vehicle in each preset driving dimension in the travel segment to be tested.

[0062] In this embodiment, the target driving data is related to the corresponding preset driving dimension. Depending on the different preset driving dimensions, there is corresponding target driving data. For example, for the speed dimension, the target driving data of the speed dimension is data related to speed, such as the speed and acceleration of the vehicle in the travel segment to be tested. For the defensive dimension, the target driving data of the defensive dimension is data related to vehicle driving safety, such as the speed of the vehicle in the travel segment to be tested, the relative position between the vehicle and other objects, and other data.

[0063] Step S250: Based on the preset dimensional evaluation model, the target driving data is processed to obtain a prediction score.

[0064] In this embodiment, for target driving data of a preset driving dimension, which includes at least one feature data, the prediction score is used to represent the driving behavior score of the feature data in the target driving data in the preset driving dimension.

[0065] For a preset driving dimension, each feature data in the target driving data corresponds to a preset dimension evaluation model. The dimension evaluation models corresponding to two feature data may be the same or different, and the dimension evaluation model corresponding to the feature data of one preset driving dimension may be the same or different from the dimension evaluation model corresponding to the feature data of another preset driving dimension.

[0066] Step S270: Obtain a target score for the trip segment to be measured based on the position of the trip segment to be measured in the mileage and the predicted score.

[0067] The target score is used to represent the driving behavior score of the vehicle in the preset driving dimension in the tested trip segment.

[0068] Of course, in some embodiments, the driving behavior score of the trip segment to be tested can also be corrected with reference to the position of the trip segment to be tested in all the mileage of the vehicle, that is, the driving behavior scores of other trip segments in the mileage of the vehicle can be referenced to obtain the target score of the trip segment to be tested more accurately.

[0069] In this embodiment, by presetting driving dimensions, driving environment category dimensions, and driving performance category dimensions, target driving data of corresponding dimensions are screened, and the target driving data is used as data of corresponding dimensions that accurately characterize the vehicle in the journey segment to be tested, thereby reducing the problem of excessive data noise during driving behavior data processing and improving the reliability of subsequent driving behavior data processing; at the same time, the feature data is processed by the dimensional evaluation model to obtain a predicted score, and the driving behavior score of the journey segment to be tested can be obtained based on the predicted score. Furthermore, the driving behavior score of the journey segment to be tested that can be obtained based on the predicted score is corrected by the position of the journey segment to be tested in the journey mileage, thereby obtaining a more accurate target score. This target score can more comprehensively reflect the driver's driving behavior preferences and driving style, and does not require manual intervention, resulting in lower costs.

[0070] For different driving dimensions, corresponding target scores can be obtained. In this way, driving behavior can be scored from multiple perspectives, further improving the accuracy of the vehicle's driving behavior analysis results.

[0071] FIG3 is a flow chart of a method for processing driving behavior data according to another embodiment. FIG3 is an implementation of step S230 in FIG2 . FIG3 includes steps S310 to S330 , which are described in detail as follows:

[0072] Step S310: Acquire acceleration data of the vehicle in the travel segment to be measured.

[0073] In this embodiment, the preset driving dimensions include driving performance dimensions, which include stability dimensions. The stability dimension generally scores driving behavior based on vehicle ride comfort. This allows for the acquisition of vehicle acceleration data. Excessive acceleration values ​​indicate reduced vehicle comfort, thereby generating target driving data.

[0074] In some embodiments, the acceleration data may include positive longitudinal acceleration, negative longitudinal acceleration and lateral acceleration; wherein the longitudinal direction is regarded as the direction of travel of the vehicle, the lateral direction is the direction orthogonal to the longitudinal direction on the travel plane of the vehicle, the positive longitudinal acceleration may be regarded as the acceleration of the vehicle accelerating, and the negative longitudinal acceleration may be regarded as the acceleration of the vehicle decelerating.

[0075] Step S330: Obtain the number of acceleration behaviors in the acceleration data whose value is greater than the preset acceleration value, and use the number of acceleration behaviors as the target driving data in the stability dimension.

[0076] In this embodiment, during the measured travel segment, the greater the acceleration value, the lower the vehicle's comfort, and thus the lower the stability driving behavior score. Therefore, the number of acceleration behaviors in the acceleration data with a value greater than the preset acceleration value is obtained, and this number of acceleration behaviors is used as the target driving data for the stability dimension. It is understood that this number of acceleration behaviors can include different directions, such as the longitudinal and lateral accelerations mentioned above, namely, the longitudinal acceleration behavior sub-number (including the longitudinal positive acceleration sub-number and the longitudinal negative acceleration sub-number) and the lateral acceleration behavior sub-number.

[0077] Among them, if the value in the acceleration data is greater than the preset acceleration value, the number of acceleration behaviors is increased by 1. If the values ​​in the acceleration data are greater than the preset acceleration value for a continuous period of time, it is actually regarded as only one behavior with a larger acceleration, that is, the number of acceleration behaviors is only increased by 1 during the continuous period of time. When the values ​​in the acceleration data at two moments or time periods are greater than the preset acceleration value, and the value in the acceleration data at the time or moment between the two moments or time periods is less than the preset acceleration value, the number of acceleration behaviors is increased by 2, that is, the moment or period when the value in the acceleration data is greater than the preset acceleration value is regarded as the first time. Then, on the time axis, the values ​​in the acceleration data at the moment before the first time and the moment after the first time are less than the preset acceleration value. The number of acceleration behaviors can be obtained by counting the number of times the first time occurs.

[0078] In some embodiments, the preset driving dimension is a defensive dimension. At this time, the target driving data can be obtained through the position between the vehicle and the target object and the speed of the vehicle, such as the first number, the number of first target vehicles, the driving time of the first target vehicle, the second number, and the number of target roads in which the vehicle speed in the target road in the measured travel segment is greater than the preset first value (first sub-parameter).

[0079] Among them, the target roads may include roads that require vehicles to drive at low speeds, such as intersections, zebra crossings, special area sections (such as school sections), etc. There is no specific restriction here. The first number is the number of times the vehicle overtakes on a curve, the number of first target vehicles is the number of times the vehicle is parallel to other vehicles and the number of times it follows the vehicle at a close distance in the measured journey segment, the driving time of the first target vehicle is the time the vehicle is parallel to other vehicles and the time it follows the vehicle at a close distance, and the second number is the number of times the vehicle performs aggressive lane changes.

[0080] In some embodiments, the preset driving dimension is the curve dimension. In this case, the target driving data may include the speed characteristic data of the vehicle on the curve section and the turning characteristic data of the vehicle on the curve section, such as the initial speed of the curve section, the turning speed limit, the terminal speed of the curve section, the lateral acceleration, and the proportion of time the vehicle receives the brake signal.

[0081] The turning speed limit is obtained by the following method: The turning radius is calculated by the following method: r m =WB / tan(σ)

[0082] Among them, r m is the turning radius in m (meters), WB is the wheelbase, and σ is the steering angle. This turning radius is generally the minimum turning radius, that is, the turning radius at the maximum steering angle. σ can be regarded as the maximum steering angle.

[0083] The turning speed limit can be regarded as the standard speed on the curved road section and can be obtained by the following method:

[0084] Among them, v0 is the turning speed limit, unit is km / h (kilometers per hour).

[0085] In some embodiments, the preset driving dimension is the speed dimension. In this case, the target driving data may include speed characteristic data characterizing the speed capability of the target vehicle and acceleration characteristic data characterizing the acceleration capability of the target vehicle, such as average vehicle speed, acceleration and maximum vehicle speed, wherein the acceleration may include static acceleration and overtaking acceleration.

[0086] In some embodiments, the preset driving dimension is the motion dimension. In this case, the target driving data may characterize parameters of the vehicle's motion performance in the travel segment to be tested, such as acceleration data and overtaking data. The acceleration data may be the number of times the vehicle's acceleration is greater than a preset acceleration threshold, and the overtaking data may include the number of overtaking on a curve and the number of aggressive lane changes.

[0087] In some embodiments, the preset driving dimension is the braking dimension. In this case, the target driving data may be data when the vehicle is in a braking state, such as braking duration, number of braking times, deceleration acceleration, and anti-lock braking times.

[0088] According to the numerical value of the deceleration acceleration in the braking state, the braking state can be divided into a first braking state and a second braking state. The numerical value of the deceleration acceleration in the first braking state is not greater than the acceleration threshold, and the numerical value of the deceleration acceleration in the second braking state is greater than the acceleration threshold. In this way, the first braking state can also be regarded as a normal braking state, and the second braking state as an emergency braking state.

[0089] The acceleration threshold can be determined by empirical parameters. For example, in some embodiments, the acceleration threshold is defined as 0.7g, and the deceleration acceleration in [-0.7g, 0) is normal braking, and the deceleration acceleration in (-∞, -0.7g) is emergency braking. The deceleration acceleration in [-0.7g, 0) is normal braking, and the deceleration acceleration in (-∞, -0.7g) should be understood as emergency braking, which is a braking state in which the maximum deceleration acceleration is in the range of [-0.7g, 0) for normal braking, and the maximum deceleration acceleration is in the range of (-∞, -0.7g) for emergency braking. Of course, 0.7g is an exemplary value, and other embodiments may also use other values.

[0090] The number of braking times may include a first number of braking times corresponding to the first braking state and a second number of braking times corresponding to the second braking state, and the deceleration acceleration includes a first deceleration acceleration in the first braking state and a second deceleration acceleration in the second braking state.

[0091] The following describes scenarios such as overtaking, aggressive lane changes, and curved roads.

[0092] In one embodiment, as shown in FIG4 , a certain range area around the vehicle is set as an overtaking detection area. For example, the size of the overtaking detection area is 10 meters x 10 meters (of course, it can also be other ranges, which is not limited here). Other vehicles around are detected at certain time intervals. If there is a vehicle B that stays in the overtaking detection area for a time greater than a preset time length, it is determined that the vehicle is traveling parallel to vehicle B. The parallel driving time of vehicle B is the time that vehicle B stays in the overtaking detection area after it is determined to be traveling parallel to the vehicle, that is, the time that vehicle B stays in the overtaking detection area minus the preset time length, that is, the driving time when the first target vehicle is vehicle B.

[0093] In some embodiments, as shown in FIG5 , if the distance between the vehicle C and the vehicle C on the y-coordinate is less than the preset longitudinal distance and the lateral distance between the vehicle C and the vehicle C on the x-coordinate is less than the preset lateral distance within a preset time period, then the vehicle C may be used as the first target vehicle whose distance from the vehicle C is less than the preset distance threshold, and the vehicle C and the vehicle C follow closely, that is, the preset distance threshold includes the preset longitudinal distance and the preset lateral distance.

[0094] The y coordinate is the direction of travel of the vehicle, the x coordinate is the direction orthogonal to the y coordinate in the plane on which the vehicle is traveling, and the distance between the vehicle and any other vehicle can be obtained by the distance between the centers of the two vehicles.

[0095] It can be understood that vehicle C should be located in front of the vehicle in the driving direction, and the distance between vehicle C and the vehicle on the y coordinate is the smallest compared with other vehicles located in front of the vehicle.

[0096] In this embodiment, the preset lateral distance and the preset longitudinal distance can be obtained through empirical parameters. The preset lateral distance is used to characterize the distance within which the vehicle and vehicle C may collide, and can be determined according to the width of the vehicle. For example, the preset lateral distance can be a value such as 2m (meters), 2.5m, etc. The preset longitudinal distance can be determined by the value of the vehicle speed. The greater the speed, the greater the preset longitudinal distance, and the two are linearly related. The preset longitudinal distance is the warning distance maintained between the vehicle and vehicle C, and can be the vehicle speed multiplied by a certain reflection time, such as the reflection time is 0.1s (seconds), 0.3s, etc., to obtain the preset longitudinal distance.

[0097] At this time, the close-following time corresponding to vehicle C is the time during which the distance between vehicle C and the first target vehicle remains less than the preset distance threshold after vehicle C is determined to be the first target vehicle, that is, the time during which the distance between vehicle C and the first target vehicle remains less than the preset distance threshold minus the preset duration.

[0098] In this way, in this embodiment, the driving time of the first target vehicle such as vehicle category B and the driving time of the first target vehicle such as vehicle category C can be obtained. The number of the first target vehicles is multiple, that is, the driving time of multiple first target vehicles and the number of first target vehicles are obtained.

[0099] In one embodiment, as shown in FIG6 , other surrounding vehicles are detected at certain time intervals. If vehicle A is detected entering the overtaking detection zone from the positive direction of the overtaking detection zone (i.e., in front of the vehicle, i.e., in the direction of travel of the vehicle) (i.e., within a continuous time, vehicle A enters the front of the overtaking detection zone from outside the overtaking detection zone), and the distance between the vehicle A and vehicle A on the x-coordinate is less than or equal to a preset coordinate value, and the distance between the vehicle A and vehicle A on the y-coordinate is greater than 0 and less than or equal to a preset coordinate value, and the speed of vehicle A is greater than the speed of the overtaken vehicle, then the overtaking vehicle confirmation of vehicle A is triggered, and the identification of the other vehicle A is recorded. If there are multiple vehicles that meet the above conditions, they are recorded simultaneously to identify and distinguish different overtaking vehicles. The other vehicles shown in FIG6 in the negative direction of the vehicle are not overtaking vehicles.

[0100] The preset coordinate value can be 5m (meters), 4.5m, etc., which is related to the width of the vehicle. The overtaking speed can be regarded as the vehicle A being in a normal driving state, or a process driving state (the speed is greater than 30km / h). It can also be a speed value for confirming a smooth driving journey segment, so as to ensure that the vehicle is overtaking normally, and not because the vehicle A has a malfunction or stops causing the vehicle to overtake.

[0101] In this embodiment, although vehicle A is marked as an overtaking vehicle, there may be two situations: overtaking unsuccessful or overtaking successful. As shown in Figure 7, if the overtaking vehicle leaves the overtaking detection zone from the negative direction (the direction opposite to the driving direction of the vehicle), it indicates that the overtaking is successful; as shown in Figure 8, if the overtaking vehicle does not leave the overtaking detection zone from the negative direction (such as from the positive direction of the overtaking detection zone, or from the x-axis direction of the overtaking detection zone), it indicates that the overtaking failed.

[0102] But no matter whether the overtaking is successful or unsuccessful, vehicle A is still the overtaking vehicle.

[0103] In this embodiment, the steering angle of the vehicle's steering wheel can be used to determine whether overtaking occurs on a curve. If the steering angle is greater than a preset angle threshold (such as 5, 6, 10, etc.) and lasts for a period of time (such as 2s, 3s, 4s, etc.), it can be regarded as the vehicle entering the curve. After driving in the curve for a period of time, if the steering angle returns to positive or is less than the preset angle threshold and also lasts for a period of time (such as 2s, 3s, 4s, etc.), it is regarded as the end of the curve, thereby obtaining the curved road section.

[0104] In this embodiment, the representative vehicle in the aggressive lane change has completed the overtaking operation, while the overtaking vehicle that failed the overtaking operation is not judged as aggressive lane change, that is, is not counted in the second number.

[0105] After successfully overtaking, the vehicle being overtaken is used as the vehicle to be tested, and lane change detection is triggered. As shown in Figure 9, a first distance in a first direction and a second distance in a second direction between the vehicle and the vehicle to be tested are obtained within a preset time period. The first direction is the y-coordinate direction, and the second direction is regarded as the x-coordinate direction. If the first distance is less than a preset first threshold and the second distance is less than a preset second threshold, it can be regarded as the vehicle aggressively changing lanes after overtaking. By counting the number of vehicles to be tested whose first distance is less than the preset first threshold and the second distance is less than the preset second threshold, the number of aggressive lane changes of the vehicle can be obtained, thereby obtaining the second number.

[0106] Among them, the second distance is used to determine whether the vehicle and the vehicle to be tested are merging into the same lane. Therefore, the second threshold value can be predicted through empirical parameters. The second distance can be the same as the preset lateral distance. If the second distance is less than the preset second threshold value, the vehicle is deemed to have merged into the same lane as the vehicle to be tested. The first distance is the distance between the vehicle and the vehicle to be tested. When the first distance is small within a preset time period, it proves that the vehicle has made an aggressive lane change. The preset time period is a certain period of time after a successful overtaking, such as within 10 seconds or 8 seconds after a successful overtaking. If the vehicle merges into the same lane with the vehicle to be tested within this shorter preset time period and the first distance is small, it is deemed to be an aggressive lane change. The preset first threshold value can also be obtained through empirical parameters, such as the speed of the vehicle to be tested. The greater the speed, the larger the preset first threshold value, and the two are linearly related.

[0107] The state of receiving the throttle input signal when the vehicle speed is less than a certain value (such as 4, 5, etc.) is regarded as the stationary acceleration state, and the state of the vehicle within a preset first time period after the stationary state is regarded as the stationary acceleration state, and the stationary acceleration is the acceleration obtained in the stationary acceleration state.

[0108] FIG10 is a flow chart of a method for processing driving behavior data according to another embodiment. FIG10 is an implementation of step S250 in FIG2 , including steps S1010 to S1030:

[0109] Step S1010: Based on the acceleration directions corresponding to the acceleration behavior times, sub-times of acceleration behaviors with different acceleration directions are filtered from the acceleration behavior times.

[0110] The preset driving dimensions include dimensions of driving performance categories, the dimensions of driving performance categories include stability dimensions, the target driving data include the number of acceleration behaviors in which the value of the vehicle's acceleration data is greater than the preset acceleration value, the acceleration direction includes longitudinal and lateral (such as lateral acceleration in a curve), and the acceleration behavior sub-number includes longitudinal acceleration behavior sub-number and lateral acceleration behavior sub-number.

[0111] S1030: According to the preset dimensional evaluation model for different acceleration directions, the acceleration behavior sub-numbers corresponding to the acceleration direction are processed to obtain a prediction score for the corresponding acceleration direction.

[0112] In this embodiment, the characteristic data of different acceleration directions correspond to different dimensional evaluation models, that is, the dimensional evaluation model corresponding to the longitudinal acceleration behavior sub-number is different from the dimensional evaluation model corresponding to the lateral acceleration behavior sub-number.

[0113] In one embodiment, the dimensionality evaluation model may be as follows, where the first hyperparameter and the second hyperparameter are different for different acceleration directions, so that the dimensionality evaluation model corresponding to the longitudinal acceleration behavior sub-order is different from the dimensionality evaluation model corresponding to the lateral acceleration behavior sub-order: μ s =a s +b s *n s

[0114] Wherein, μ is the prediction score, a is the first hyperparameter, b is the second hyperparameter, n is the number of acceleration behaviors, and different acceleration directions correspond to the first hyperparameter and the second hyperparameter. s is the sequence number of the acceleration direction. In one embodiment, the sequence number corresponding to the horizontal direction is 1, and the sequence number corresponding to the vertical direction is 2. s Can be 0.

[0115] In some embodiments, b s 、a s The value range is [0,1], such as b s It can be 0, μ1=0.1, μ2=0.05, that is, the prediction score of the longitudinal feature data is 0.05, and the prediction score of the horizontal feature data is 0.1.

[0116] In some embodiments, the preset driving dimension is the defensive dimension, and the corresponding dimension evaluation model is: m =c m +d m *f m

[0117] Among them, u m is the prediction score, c m is the third hyperparameter, d m is the fourth hyperparameter, f mis the target driving data, m is the mth feature data in the target driving data, and the range of m is [1,5].

[0118] The third hyperparameter and the fourth hyperparameter corresponding to different feature data may be the same or different, and are obtained through empirical parameters.

[0119] In some embodiments, c m d m The value range is [0,1], such as d m is 0, then the prediction score for different feature data in the target driving data is u m =c m .

[0120] In this embodiment, u1=0.05, u1 is the first prediction score corresponding to the first sub-parameter, u2=0.05, u2 is the second prediction score corresponding to the first quantity, u3=0.05, u3 is the third prediction score corresponding to the second quantity, u4=0.05, u4 is the fourth prediction score corresponding to the number of the first target vehicles, and u5=0.0005, u5 is the fifth prediction score corresponding to the driving time of the first target vehicle.

[0121] In some embodiments, the preset driving dimension is the curve dimension, and the corresponding dimension evaluation model is: f = (Min(1,v1 / v0)*L*∈1+Min(1,v2 / v0)*L*∈2+Min(1,a m / a0)*L*∈3)*(1-k t )

[0122] Among them, f is the prediction score, v1 is the initial speed of the curved road section, v0 is the turning speed limit, v2 is the terminal speed of the curved road section, and a m is the lateral acceleration, k t is the proportion of time the vehicle receives the brake signal, L is the preset scoring threshold, ∈1, ∈2, ∈3 are weights, and their value range is (0, 1). In other embodiments, ∈1 = 0.2, ∈2 = 0.2, ∈3 = 0.6, and a0 is a constant parameter of lateral acceleration.

[0123] In this embodiment, the value range of the prediction score is [0, L], where L is a preset scoring threshold, i.e., the maximum value that can be obtained for the driving behavior score. The specific value can be set by the user, and the value range is [1, 100], such as 5, 10, 100, etc.

[0124] In this embodiment, the travel segment to be measured includes multiple curved road segments, and the data of one curved road segment is used as a feature data, that is, the predicted score is the driving behavior score of each curved road segment.

[0125] In some embodiments, the preset driving dimension is the speed dimension, and the characteristic data of the target driving data corresponds to the evaluation model of each dimension. For example, the dimension evaluation model corresponding to the average vehicle speed is: f1(V0)=Max(V0-V q ,0) / (V p -V q )*L

[0126] Among them, f1(V0) is the prediction score corresponding to the average vehicle speed, V0 is the average vehicle speed, V q 、V p is a constant parameter of the average speed.

[0127] In this embodiment, V q It can be regarded as a certain speed value for entering a smooth driving section, and its value range is [20-40]. For example, any value between 30 and 40, V q The speed limit value of some road sections can be determined according to the specific speed limit value of the smooth driving section in the test section. Its value range is [60,120], such as 60, 90, etc. The dimensional evaluation model corresponding to the maximum speed is: f4(V m )=1+L*Max(0,(V m -V k )) / (100*Max(0,(V m -V k ))+3000)

[0128] Among them, f4(V m ) is the prediction score corresponding to the maximum vehicle speed, V m is the maximum vehicle speed, V k It is a constant parameter of the maximum vehicle speed, and its value range is [60,120]. For example, it can be 60, 90, etc.

[0129] The dimensional evaluation model corresponding to the acceleration feature data is: f2(a(m))=a(m) / g0*L

[0130] Among them, f2(a(m)) is the prediction score corresponding to the acceleration feature data, a(m) is the acceleration feature data, g0 is the normalized acceleration threshold constant, g0 is a multiple of g, and its value range is [0.3g, 0.7g], such as 0.5g, 0.4g, etc., and g is the vehicle acceleration measurement unit, which is a constant.

[0131] Through the evaluation model corresponding to the acceleration characteristic data, the predicted scores of the stationary acceleration and the overtaking acceleration can be obtained respectively. That is, the predicted scores corresponding to the acceleration characteristic data include: the predicted score of the stationary acceleration and the predicted score of the overtaking acceleration.

[0132] For example, the predicted score of static acceleration is set to f2(a1(m)), and the predicted score of overtaking acceleration is set to f 22 (a2(m)), then the evaluation model corresponding to the acceleration characteristic data of the static acceleration a1(m) and the overtaking acceleration a2(m) can be obtained, that is: f2(a1(m))=a1(m) / g0*L f2(a2(m))=a1(m) / g0*L

[0133] Where g is the customary unit of vehicle acceleration.

[0134] In some embodiments, the preset driving dimension is the motion dimension, and the characteristic data of the target driving data corresponds to the evaluation model of each dimension, such as the motion evaluation function of the acceleration sub-parameter corresponding to the characteristic data is: f a (a)=Min(1,a / p1)*L

[0135] Among them, f a (a) is the prediction score corresponding to the acceleration sub-parameter, p1 is a hyperparameter, and its value range can be [90, 200], such as 100, and a is the acceleration sub-parameter.

[0136] The dimensional evaluation function corresponding to the number of overtaking on a curve is: f b (b)=Min(1,b / p2)*L

[0137] Among them, f b (b) is the predicted score corresponding to the number of overtaking times, p2 is a hyperparameter, and its value range can be [500, 2000], such as 1000, and b is the number of overtaking times on the curve.

[0138] The dimension evaluation function corresponding to the number of sports overtaking is: f c (c) = a + b * c

[0139] Among them, f c (c) is the predicted score corresponding to the number of sports overtaking, c is the number of sports overtaking, and its value is the value obtained by adding the number of curve overtaking and the number of aggressive lane changes. a and b are hyperparameters, and their value range can be [0,1]. For example, b can be 0. In some embodiments, f c (c) = a, where a is 0.05.

[0140] In some embodiments, the preset driving dimension is the braking dimension, and the characteristic data of the target driving data are divided into different categories. The characteristic data of different categories correspond to the evaluation models of each dimension. For example, the characteristic data of the braking time category includes the braking duration and the mechanical braking duration, the characteristic data of the braking state category includes the first deceleration acceleration and the first braking number of the first braking state, the second deceleration acceleration and the second braking number of the second braking state, the characteristic data of the anti-lock braking category includes the anti-lock braking number, and the characteristic data of the braking number category includes the braking number.

[0141] In some embodiments, the dimension evaluation model of the braking times category of the feature data of the braking time category is: f1(T0,T m )=(T0-T m ) / T0*L

[0142] Among them, f1(T0,T m ) is the predicted score of the braking time category, T0 is the braking time, T m is the mechanical braking duration.

[0143] In some embodiments, the dimensional evaluation model of the braking status category is:

[0144] f2(a(m),a′(n))=m / (m+n)*(s1 / s2*a(m)+s3)+n / (m+n)*(-s4*a′(n)-s5)

[0145] Among them, f2(a(m1),a′(n)) is the predicted score of the braking state category, m is the first braking number, n is the second braking number, a(m1) is the first deceleration acceleration, a′(n) is the second deceleration acceleration, s1, s2, s3, s4, and s5 are fitting constants of the state braking evaluation model, which can be obtained by fitting according to empirical parameters.

[0146] In some embodiments, the values ​​of the fitting constants of the state braking assessment model may be the same or different. In other embodiments, the range is [0, 100], such as s1=50, s2=7, s3=5, s4=10, s5=7, etc.

[0147] The dimensional evaluation model of the anti-lock braking category is: f4(N a )=l+N a *k

[0148] Among them, N a is the number of anti-lock braking times, f4(N a ) is the prediction score of the anti-lock braking class, l and k are hyperparameters, and their value range is [0,1]. k can be 0. In some embodiments, k=0 and l is 0.1.

[0149] The dimensional evaluation model of the braking number category is: f3(N b )=d1*N b +d2

[0150] Among them, N b is the number of braking times, f3(N b ) is the predicted score of the braking number category, d1 and d2 are the fitting constants of the braking number evaluation model, which can be obtained by fitting according to empirical parameters.

[0151] In some embodiments, d1 and d2 can be positive or negative, and can be the same or different. In other embodiments, the range of d1 and d2 is [-10, 20]. For example, d1 can be -0.015, 0.03, etc., and d2 can be 4.5, 6, etc. There is no specific limitation here.

[0152] The driving behavior data processing device provided in this embodiment can be used to execute the above-mentioned driving behavior data processing method. Its implementation principle and technical effects are similar and will not be described in detail in this embodiment.

[0153] FIG11 is a flow chart of a method for processing driving behavior data according to another embodiment. FIG11 is an implementation of step S270 in FIG2 , including steps S1110 to S1150:

[0154] Step S1110: Obtain the initial travel segment.

[0155] Among them, the initial trip segment is located before the trip segment to be measured in the vehicle's mileage; that is, the vehicle's mileage can be divided into multiple trip segments, and the initial trip segment is located before the trip segment to be measured. To ensure the uniformity of the data volume in the trip segments, the multiple trip segments are divided in the same way. For example, if divided by time, the driving time of each trip segment is the same. If divided by distance, the travel distance of each trip segment is the same.

[0156] It should be noted that the trip segment mentioned here is the trip segment where driving behavior data processing begins. That is, although the vehicle's actual mileage is very long, the mileage for which driving behavior data processing has not been performed to obtain the target score for the corresponding trip segment is not taken into consideration. That is, the trip mileage is the trip for which driving behavior data processing is performed.

[0157] Step S1130: If the initial trip segment does not exist, a first score is obtained based on the predicted score, and the first score is used as the target score.

[0158] In this embodiment, the initial travel segment does not exist, and the travel segment to be measured is the first travel segment for driving behavior data processing, that is, the travel mileage mentioned above starts from the travel segment to be measured.

[0159] In one embodiment, step S1130 may include S10 to S11: S10: obtaining a prediction score and target driving data; S11: obtaining a first score based on the prediction score and target driving data based on a preset target evaluation model.

[0160] For the stability dimension, the target driving data includes the number of acceleration behaviors in different acceleration directions, and the predicted scores include the predicted scores of the number of acceleration behaviors in different acceleration directions. The acceleration directions include longitudinal and lateral directions. The target evaluation model is: Score s =Max(0,(L-n1*μ1-n2*μ2))*δ

[0161] Among them, Score s is the first score of the stability dimension, n1 is the number of acceleration behaviors in the longitudinal positive acceleration direction, n2 is the number of acceleration behaviors in the longitudinal negative acceleration direction, n3 is the number of acceleration behaviors in the lateral acceleration direction, δ is a preset weighted value, μ1 is the predicted score of the longitudinal acceleration behavior sub-number, μ2 is the predicted score of the lateral acceleration behavior sub-number, in one embodiment, the value range of δ is [0.2, 0.8], and δ can be a value such as 0.6 or 0.5.

[0162] For the defensive dimension, the target evaluation model is: Score d =(L-u1*n1-u2*n2-μ3*n3-μ4*n4--μ5*n5)*δ

[0163] Among them, Score d is the first score of the defensive dimension, n1 is the first sub-parameter, u1 is the first prediction score, n2 is the first number, u2 is the second prediction score, n3 is the second number, u3 is the third prediction score, n4 is the number of the first target vehicles, u4 is the fourth prediction score, n5 is the driving time of the first target vehicle, in seconds, u5 is the fifth prediction score, and δ is a preset weighted value.

[0164] For the curve dimension, the target evaluation model is: Score c =f1*δ

[0165] Among them, Score c is the first score of the curve dimension, and f1 is the median or median average or average value of the predicted scores of all curve road sections in the trip segment to be tested.

[0166] For the speed dimension, the target evaluation model is: Score h =(Min(L,f1(V0))*γ1+Min(L,f2(a1(m)))*γ2+Min(L,f2(a2(m)))*γ3)*f4(Vm )*δ

[0167] Among them, Score h is the first score of the speed dimension, f1(V0) is the predicted score corresponding to the average vehicle speed, f2(a(m)) is the predicted score corresponding to the acceleration feature data, and f4(V m ) is the predicted score corresponding to the maximum vehicle speed, γ1, γ2, and γ3 are weights. In one embodiment, δ can be a value such as 0.6 or 0.5.

[0168] In some embodiments, the sum of γ1, γ2, and γ3 is 1, and the three weights are all decimals with a value range of (0, 1). The specific setting of the weight can refer to empirical parameters. For example, γ1 corresponds to the weight of f1(V0), which is determined according to the importance of the average vehicle speed in the speed dimension to the driving behavior score. In other embodiments, γ1 = 0.4, γ2 = 0.3, and γ3 = 0.3.

[0169] For the motion dimension, the target evaluation model is: Score a =(f a (a)*β1+f b (b)*β2+f c (c)*c)*δ

[0170] Among them, Score a is the first fraction of the motion dimension, f a (a) is the prediction score corresponding to the accelerometer parameter, f b (b) is the prediction score corresponding to the number of overtaking times, c is the number of sports overtaking times, and f c (c) is the prediction score corresponding to the number of sports overtaking, and β1 and β2 are weights.

[0171] Since the acceleration sub-parameters include positive acceleration characteristic data and negative acceleration characteristic data, f a The weight of (a) corresponds to the predicted score corresponding to the positive acceleration feature data and the predicted score corresponding to the negative acceleration feature data, that is, the sum of β1 and 2β2 is 1, and the two weights are both decimals, and their value range is (0,1). The specific setting of the weight can refer to the empirical parameters. In other embodiments, β1 = 0.4, β2 = 0.3.

[0172] For the braking dimension, the target evaluation model is: Score b =(f1(T0,T m )*α1+f2(a(m),a′(n))*α2+f3(N b )*α3-α4*f4(N a ))*δ

[0173] Among them, Scoreb is the first fraction of the braking dimension, f1(T0,T m ) is the prediction score of the braking time category, f2(a(m1),a′(n)) is the prediction score of the braking state category, f4(N a ) is the prediction score of the anti-lock braking class, f3(N b ) is the predicted score of the braking number category, N b is the number of braking times, N a is the number of anti-lock braking times, and α1, α2, and α3 are weights.

[0174] In some embodiments, the sum of α1, α2, α3 and α4 is 1, and the four weights are all decimals, and their value range is (0,1). The specific setting of the weight can refer to the empirical parameters, such as α1 corresponds to f1(T0,T m ) is determined according to the importance of the characteristic data in the braking time category to the driving behavior score. In other embodiments, α1 = 0.3, α2 = 0.4, α3 = 0.2, and α4 = 0.1.

[0175] Step S1150: If the initial trip segment exists, obtain a second score based on the predicted score, and obtain a target score based on the difference between the third score of the initial trip segment and the second score.

[0176] The ratio between the first score and the second score is a preset weighted value.

[0177] In some embodiments, the third score of the initial trip segment is obtained in the same manner as the target score of the trip segment to be tested, both representing the driving behavior score of the initial trip segment.

[0178] In this embodiment, the value of the second score is the value of the first score divided by the preset weighted value.

[0179] In some embodiments, step S1150 may include S20 to S21: S20: If the difference between the third score and the second score is within a preset numerical range, the second score evaluation score is used as the target driving behavior evaluation score; S21: If the difference between the third score and the second score is not within the preset numerical range, the target score is obtained based on the third score and the preset error value.

[0180] In this embodiment, the difference between the third score and the second score is the value obtained by subtracting the third score from the second score. The preset error interval can be obtained through empirical parameters. For example, the preset error interval is [-0.5, 0.5]. The fixed error value is the extreme value of the preset error interval. When calculating the target score, the extreme value to which the difference is close is used as the fixed error value for the final calculated target score. If the difference is 0.6, the third score is added with 0.5 to obtain the target score.

[0181] The numerical value of the control target score is in the range of [0, L].

[0182] In this embodiment, a driving behavior score for the trip segment to be tested is comprehensively obtained based on its position in the trip mileage. Thus, the obtained score can avoid the problem of low target score accuracy caused by data errors in the trip segment to be tested. When calculating the target score for the trip segment to be tested, the driving behavior scores of corresponding driving dimensions of other trip segments are comprehensively considered to obtain a target score that accurately reflects the driver's driving behavior preferences and driving style.

[0183] 12 is a structural diagram of a driving behavior data processing device according to an embodiment. The driving behavior data processing device 5 includes: a to-be-tested trip segment determination module 510, configured to obtain the to-be-tested trip segment based on the vehicle's mileage; a target driving data acquisition module 530, configured to filter target driving data corresponding to a preset driving dimension from the driving data of the to-be-tested trip segment based on a preset driving dimension; wherein the preset driving dimension includes a dimension of a driving environment category and / or a dimension of a driving performance category; a predicted score acquisition module 550, configured to process the target driving data based on a preset dimensional evaluation model to obtain a predicted score; wherein the predicted score is used to characterize the driving behavior score of the feature data in the target driving data in the preset driving dimension; and a target score acquisition module 570, configured to obtain a target score for the to-be-tested trip segment based on the position of the to-be-tested trip segment in the mileage and the predicted score, wherein the target score is used to characterize the driving behavior of the vehicle in the preset driving dimension in the to-be-tested trip segment.

[0184] In one implementable manner, the preset driving dimension includes the dimension of the driving performance category, and the dimension of the driving performance category includes the stability dimension; the target driving data acquisition module 530 includes: an acceleration acquisition unit, configured to obtain the acceleration data of the vehicle in the travel segment to be tested; a target driving data acquisition unit, configured to obtain the number of acceleration behaviors in the acceleration data whose value is greater than the preset acceleration value, and use the number of acceleration behaviors as the target driving data on the stability dimension.

[0185] In one implementable manner, the preset driving dimension includes the dimension of the driving performance category, the dimension of the driving performance category includes the stability dimension, and the target driving data includes the number of acceleration behaviors in which the value of the vehicle's acceleration data is greater than the preset acceleration value; the prediction score acquisition module 550 includes: an acceleration screening unit, configured to screen the acceleration behavior sub-numbers of different acceleration directions in the acceleration behavior number based on the acceleration direction corresponding to the acceleration behavior number; a prediction score acquisition unit, configured to process the acceleration behavior sub-numbers of the corresponding acceleration direction according to the preset dimensional evaluation model for different acceleration directions to obtain the prediction score of the corresponding acceleration direction.

[0186] In one implementation, the dimension evaluation model preset for different acceleration directions in the prediction score acquisition unit is: μ x =a x +b x *n x

[0187] Where μ is the prediction score, a is the first hyperparameter, b is the second hyperparameter, n is the number of acceleration behaviors, different acceleration directions correspond to the first and second hyperparameters, and x is the sequence number of the acceleration direction.

[0188] In one implementation, the target score acquisition module includes: an initial trip segment acquisition unit, configured to acquire the initial trip segment; wherein the initial trip segment is located before the trip segment to be measured in the vehicle's mileage; a first target score acquisition unit, configured to acquire a first score based on the predicted score if the initial trip segment does not exist, and use the first score as the target score; a second target score acquisition unit, configured to acquire a second score based on the predicted score if the initial trip segment exists, and acquire the target score based on the difference between the third score of the initial trip segment and the second score, wherein the ratio between the first score and the second score is a preset weighted value.

[0189] In one implementation, the preset driving dimension includes a dimension of a driving performance category, and the dimension of the driving performance category includes a stability dimension; the first target score acquisition unit includes: a data acquisition module, configured to acquire a predicted score and target driving data; wherein the target driving data includes acceleration behavior sub-numbers in different acceleration directions, the predicted score includes the predicted scores of acceleration behavior sub-numbers in different acceleration directions, and the acceleration direction includes longitudinal and lateral directions; the first target score acquisition module is configured to obtain a first score based on the predicted score and the target driving data based on a preset target evaluation model; wherein the target evaluation model is: Score s =Max(0,(L-n1*μ1-n2*μ2))*δ

[0190] Among them, Score s is the first score, n1 is the number of acceleration behaviors in the longitudinal positive acceleration direction, n2 is the number of acceleration behaviors in the longitudinal negative acceleration direction, n3 is the number of acceleration behaviors in the lateral acceleration direction, δ is the preset weighted value, μ1 is the predicted score of the longitudinal acceleration behavior sub-number, μ2 is the predicted score of the lateral acceleration behavior sub-number, and L is the preset scoring threshold.

[0191] In one implementation, the second target score acquisition includes: a second target score acquisition module, configured to use the second score evaluation score as the target driving behavior evaluation score if the difference between the third score and the second score is within a preset numerical range; and a third target score acquisition module, configured to obtain the target score based on the third score and a preset error value if the difference between the third score and the second score is not within the preset numerical range.

[0192] The driving behavior data processing device provided in this embodiment can be used to execute the above-mentioned driving behavior data processing method. Its implementation principle and technical effects are similar and will not be described in detail in this embodiment.

[0193] The present disclosure further provides a vehicle, comprising the driving behavior data processing device 5 as proposed above. The functions or modules in the driving behavior data processing device 5 can be used to execute the driving behavior data processing method proposed above.

[0194] Figure 13 is a block diagram of an electronic device according to an exemplary embodiment. Please refer to Figure 13. The electronic device 1300 may include: a processor 131 and a memory 132, wherein the processor 131 and the memory 132 can communicate; exemplarily, the processor 131 and the memory 132 communicate via a communication bus 133, the memory 132 is used to store instructions, and the processor 131 is used to call the instructions in the memory to execute the driving behavior data processing method shown in any of the above method embodiments.

[0195] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), or application-specific integrated circuits (ASICs). A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the present disclosure may be directly implemented by a hardware processor or implemented by a combination of hardware and software modules in the processor.

[0196] The present disclosure provides a computer-readable storage medium having computer-executable instructions stored thereon; when the computer-executable instructions are executed by a processor, they are used to implement the driving behavior data processing method as described in any of the above embodiments.

[0197] An embodiment of the present disclosure provides a computer program product, which includes a computer program. When the computer program is executed, it enables a computer to execute the above-mentioned driving behavior data processing method.

[0198] Other embodiments of the present disclosure will readily occur to those skilled in the art after considering the specification and practicing the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the present disclosure being indicated by the following claims.

[0199] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method for processing driving behavior data, wherein: include: Obtaining a travel segment to be measured based on the vehicle's mileage; Based on preset driving dimensions, target driving data corresponding to the preset driving dimensions are screened from the driving data of the travel segment to be measured; wherein the preset driving dimensions include dimensions of a driving environment category and / or dimensions of a driving performance category; Based on a preset dimensional evaluation model, the target driving data is processed to obtain a prediction score; wherein the prediction score is used to represent the driving behavior score of the feature data in the target driving data in the preset driving dimension; Based on the position of the trip segment to be measured in the mileage and the predicted score, a target score for the trip segment to be measured is obtained, wherein the target score is used to represent a driving behavior score of the vehicle in a preset driving dimension in the trip segment to be measured.

2. The method according to claim 1, wherein The preset driving dimensions include dimensions of a driving performance category, and the dimensions of the driving performance category include a stability dimension; The step of filtering target driving data corresponding to the preset driving dimension from the driving data of the travel segment to be measured based on the preset driving dimension includes: Acquiring acceleration data of the vehicle in the travel segment to be measured; The number of acceleration behaviors whose values ​​are greater than a preset acceleration value in the acceleration data is obtained, and the number of acceleration behaviors is used as the target driving data in the stability dimension.

3. The method according to claim 1 or 2, wherein: The preset driving dimension includes a dimension of a driving performance category, the dimension of the driving performance category includes a stability dimension, and the target driving data includes the number of acceleration behaviors in which the value of the acceleration data of the vehicle is greater than a preset acceleration value; The target driving data is processed based on a preset dimensional evaluation model to obtain a prediction score, including: Based on the acceleration directions corresponding to the acceleration behavior times, filtering acceleration behavior sub-times with different acceleration directions from the acceleration behavior times; According to the preset dimensional evaluation model for different acceleration directions, the acceleration behavior sub-numbers of the corresponding acceleration directions are processed to obtain the prediction score of the corresponding acceleration direction.

4. The method according to claim 3, wherein: The preset dimensional evaluation model for different acceleration directions is: μ x =a x +b x *n x Where μ is the prediction score, a is the first hyperparameter, b is the second hyperparameter, n is the number of acceleration behaviors, different acceleration directions correspond to the first and second hyperparameters, and x is the sequence number of the acceleration direction.

5. The method according to any one of claims 1 to 4, wherein The step of obtaining a target score for the trip segment to be measured based on the position of the trip segment to be measured in the mileage and the predicted score includes: Acquire an initial travel segment; wherein the initial travel segment is located before the travel segment to be measured in the mileage of the vehicle; If the initial trip segment does not exist, obtaining a first score based on the predicted score, and using the first score as the target score; If the initial trip segment exists, a second score is obtained based on the predicted score, and the target score is obtained based on a difference between a third score of the initial trip segment and the second score, wherein a ratio between the first score and the second score is a preset weighted value.

6. The method according to claim 5, wherein: The preset driving dimension includes a dimension of a driving performance category, and the dimension of the driving performance category includes a stability dimension; and obtaining a first score based on the predicted score includes: Obtaining the predicted score and target driving data; wherein the target driving data includes acceleration behavior sub-numbers in different acceleration directions, and the predicted score includes the predicted scores of the acceleration behavior sub-numbers in different acceleration directions, and the acceleration directions include longitudinal and lateral directions; Obtaining the first score based on the predicted score and the target driving data based on a preset target evaluation model; Among them, the target evaluation model is: Score s =Max(0,(L-n1*μ1-n2*μ2))*δ Among them, Score s is the first score, n1 is the number of acceleration behaviors in the longitudinal positive acceleration direction, n2 is the number of acceleration behaviors in the longitudinal negative acceleration direction, n3 is the number of acceleration behaviors in the lateral acceleration direction, δ is the preset weighted value, μ1 is the predicted score of the longitudinal acceleration behavior sub-number, μ2 is the predicted score of the lateral acceleration behavior sub-number, and L is the preset scoring threshold.

7. The method according to claim 5 or 6, wherein: The obtaining the target score based on the difference between the third score of the initial trip segment and the second score includes: If the difference between the third score and the second score is within a preset value range, the second score evaluation score is used as the target driving behavior evaluation score; If the difference between the third score and the second score is not within a preset value range, the target score is obtained based on the third score and a preset error value.

8. A driving behavior data processing device, wherein: include: a module for determining a travel segment to be measured, configured to obtain the travel segment to be measured based on the mileage of the vehicle; a target driving data acquisition module configured to filter target driving data corresponding to a preset driving dimension from the driving data of the travel segment to be measured based on a preset driving dimension; wherein the preset driving dimension includes a dimension of a driving environment category and / or a dimension of a driving performance category; a prediction score acquisition module configured to process the target driving data based on a preset dimensional evaluation model to obtain a prediction score; wherein the prediction score is used to represent the driving behavior score of the feature data in the target driving data in the preset driving dimension; A target score acquisition module is configured to acquire a target score for the trip segment to be measured based on the position of the trip segment to be measured in the mileage and the predicted score, wherein the target score is used to characterize the driving behavior of the vehicle in a preset driving dimension in the trip segment to be measured.

9. An electronic device, wherein: The electronic device includes: a processor and a memory; the memory is used to store instructions; the processor is used to execute the instructions in the memory, so that the electronic device executes the driving behavior data processing method according to any one of claims 1 to 7.

10. A computer-readable storage medium, wherein: The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the driving behavior data processing method according to any one of claims 1 to 7.

11. A computer program product, wherein: The method comprises a computer program, which, when executed by a processor, implements the driving behavior data processing method according to any one of claims 1 to 7.

12. A vehicle, wherein: Comprising the driving behavior data processing device as claimed in claim 8.

Images