JP7302411B2 - Driving skill evaluation device - Google Patents

Description

The present invention relates to a driving skill evaluation device for evaluating the driving skill of a vehicle driver.

A driving skill evaluation device disclosed in Patent Document 1 (hereinafter referred to as a "conventional device") stores a history of driving operations in a travel section in which the driving scene of a vehicle is determined to be a normal scene for evaluation of driving skill. is stored as a history of driving operations. The conventional device determines that the driving scene of the vehicle is a steady scene (a situation where the vehicle is traveling on a straight line or a curve and there are no other vehicles around), based on the road shape of the traveling section. Set the normative driving maneuver. The conventional device is based on the history of the driver's driving operations (steering, depression of the brake pedal, depression of the accelerator pedal) in the travel section and the reference operation (reference steering, reference acceleration/deceleration operation) of the travel section. , to evaluate the driving skills of the driver.

JP 2015-219830 A

The difference in driving operation due to the driving skill of the driver tends to become noticeable when the vehicle speed is high and the vehicle is traveling in a curve section with a small radius of curvature. On the other hand, since the normal scene of the conventional device is not limited to such a situation, the difference in the driving operation due to the driving skill is less likely to appear when the driving skill is evaluated. Therefore, in the conventional device, there is a possibility that the accuracy of the driving skill evaluation will become low.

In addition, the difference in driving operation due to driving skill is the smoothness of steering wheel operation when the vehicle is traveling on a curve section, the steering operation time (the higher the skill, the longer the steering operation time), and the steering holding time (the higher the skill, the longer the steering operation time). The steering holding time is short) and the switchback operation time (the higher the skill, the longer the switchback operation time). On the other hand, the conventional device does not use such operation feature amounts (turning operation time, steering holding time, and steering back operation time) for driving skill evaluation. Therefore, in the conventional device, there is a possibility that the accuracy of the driving skill evaluation will become low.

Further, human operation is highly uncertain, and a driver with a certain driving characteristic (driving skill level) does not always perform a driving operation according to that driving characteristic. In contrast, conventional devices do not take measures against such uncertainty of human operation. That is, the conventional device evaluates the driving skill of the driver for each predetermined driving scene (steady scene) to determine the driving skill level of the driver. For this reason, for example, when a driver with a high driving skill level accidentally performs a driving operation that is determined to have a low driving skill level, the conventional device determines the driving skill level of the driver by the driving operation. may be judged (determined) to be low. Therefore, in the conventional device, there is a possibility that the accuracy of the driving skill evaluation will become low.

The present invention has been made to address the above-mentioned problems. That is, one of the objects of the present invention is to provide a driving skill evaluation device capable of improving the accuracy of driver's driving skill evaluation. Hereinafter, the driving skill evaluation device of the present invention may be referred to as "the driving skill evaluation device of the present invention".

The driving skill evaluation apparatus of the present invention includes detection units (11, 12, 13, 14, 15, 16, 17, 18); a storage unit (10) including a driving history database (10a) for storing specific driving operation related data among the driving operation related data acquired by the detecting unit; a driving skill evaluation unit (10) for evaluating the driving skill of the driver by determining the driving skill level of the driver, wherein the driving skill evaluation unit evaluates the vehicle at a predetermined vehicle speed or higher and at a predetermined curvature; The driving operation-related data when the vehicle travels in a curve section with a radius equal to or less than the radius is acquired as history data for evaluation (determined as "Yes" in step 510), and based on a predetermined feature amount possessed by the history data for evaluation obtained using a predetermined classification model for classifying to which of a plurality of classes corresponding to the driving skill levels divided into a plurality of stages the class to which the history data for evaluation belongs belongs The class to which the evaluation history data belongs is determined, and the evaluation history data is labeled with a label corresponding to the determined class (step 525), stored in the driving history database, and stored in the driving history database. , the driving skill level corresponding to the class corresponding to the largest number of the labels is determined as the driving skill level of the driver (step 530). , step 535).

According to the driving skill evaluation device of the present invention, the accuracy of driver's driving skill evaluation can be improved.

In the above description, in order to facilitate understanding of the present invention, names and/or symbols used in the embodiments are added in parentheses to configurations of the invention corresponding to the embodiments to be described later. However, each component of the present invention is not limited to the embodiments defined by the names and/or symbols.

FIG. 1 is a schematic configuration diagram of a driving skill evaluation device according to an embodiment of the present invention. FIG. 2 is a diagram for explaining curve information. FIG. 3 is a diagram for explaining determination conditions for a specific driving scene. FIG. 4 is a graph for explaining the standard operation time. FIG. 5 is a flow chart showing a routine executed by the CPU of the driving skill evaluation ECU.

<Configuration>
A driving skill evaluation device according to an embodiment of the present invention (hereinafter sometimes referred to as "this embodiment device") is applied to (mounted on) a vehicle.

As shown in FIG. 1, the embodiment device includes a driving skill evaluation ECU 10. FIG. The driving skill evaluation ECU 10 is an electric control unit having a microcomputer as a main part. The driving skill evaluation ECU 10 and other ECUs (not shown) are connected via a CAN (Controller Area Network) so as to be able to transmit and receive information to each other. A microcomputer includes a CPU, a ROM, a RAM, a readable and writable nonvolatile memory, an interface I/F, and the like. The CPU implements various functions by executing instructions (programs, routines) stored in the ROM.

A non-volatile memory of the driving skill evaluation ECU 10 stores a driving history database 10a.

The driving skill evaluation ECU 10 is connected to the following sensors, etc., and acquires detection signals or output signals of these sensors, etc. periodically (every predetermined period of time).

The camera sensor 11 acquires image data by photographing the scenery in front of the vehicle, and based on the road image data corresponding to the road included in the image data, the left and right sides of the road (own lane) on which the vehicle is traveling, based on the road image data corresponding to the road included in the image data. Recognize the right parcel line. The camera sensor 11 outputs information (curve information) on the road shape based on the recognition result.

The accelerator pedal operation amount sensor 12 detects the operation amount (accelerator opening) of the accelerator pedal of the vehicle and outputs a signal representing the accelerator pedal operation amount.
A brake pedal operation amount sensor 13 detects the amount of operation of the brake pedal of the vehicle and outputs a signal representing the amount of brake pedal operation.

The steering angle sensor 14 detects the steering angle, which is the rotation angle of the steering wheel of the vehicle caused by steering of the steering wheel by the driver of the vehicle, and outputs a signal representing the steering angle. The driving skill evaluation ECU 10 calculates the steering angular velocity, which is the amount of change in the steering angle per unit time.
The vehicle speed sensor 15 detects the traveling speed (vehicle speed) of the vehicle and outputs a signal representing the vehicle speed. The vehicle speed sensor 15 may be a wheel speed sensor provided for each of the four wheels of the vehicle. In this case, the driving skill evaluation ECU 10 obtains the vehicle speed indicating the speed of the vehicle based on the wheel speed of each wheel detected by the wheel speed sensor.

The acceleration sensor 16 detects acceleration of the vehicle and outputs a signal representing the detected acceleration. Note that when the acceleration is a negative value, the magnitude (absolute value) of the acceleration represents deceleration.

A yaw rate sensor 17 detects the yaw rate of the vehicle and outputs the yaw rate.

The GPS receiver 18 acquires the current position of the vehicle based on signals (radio waves) from GPS (Global Positioning System) satellites and outputs a signal representing the current position. Based on the acquired current position, the driving skill evaluation ECU 10 acquires data on the driving road from map information stored in a map database (not shown). The map information includes road information including the curvature radius or curvature of the road, which indicates the degree of curvature of the road.

<Outline of operation>
(Obtaining historical data for evaluation)
The driving skill evaluation ECU 10 periodically acquires data based on the signals output from the above-described sensors, etc. do.

The driving skill evaluation ECU 10 stores predetermined data among the data obtained when the driving scene of the vehicle is the specific driving scene as evaluation history data in the driving history database 10a.

The history data for evaluation includes curve information of a curve section when the driving scene of the vehicle is a specific driving scene (curvature radius R, curve distance D, and transition curve length C shown in FIG. 2), time-series data of the steering angle, It includes time-series data of steering angular velocity, time-series data related to vehicle motion (for example, time-series data of yaw rate and time-series data of rate of change of yaw rate, etc.). The history data for evaluation is stored in the driving history database 10a for each specific driving scene.

The specific driving scene is a vehicle driving scene in which a driver's driving characteristic difference (driving skill level difference) is likely to occur. Whether or not the driving scene of the vehicle is the specific driving scene is determined as follows. That is, when the vehicle travels in a curve section, the driving skill evaluation ECU 10 calculates the curve information (curvature radius R) of the curve section traveled by the vehicle and the average vehicle speed when traveling in the curve section (hereinafter referred to as "curve section average vehicle speed"). is called.). The curve section average vehicle speed is the average of the vehicle speeds from when the steering angular velocity becomes equal to or greater than a predetermined threshold (at the start of operation) to when the steering angular velocity becomes equal to or less than a predetermined threshold (stop of operation) (at the completion of the operation). .

Then, the driving skill evaluation ECU 10 determines that the obtained curve information (curvature radius R) and curve section average vehicle speed are predetermined values on the coordinate plane (vertical axis: curvature radius R, horizontal axis: curve section average vehicle speed) shown in FIG. When it is within the range AR1, it is determined that the driving scene of the vehicle is the specific driving scene. In this example, the predetermined range AR1 is a range in which the curvature radius R is 200 m or less and the curve section average vehicle speed is 40 km/h or more and 60 km/h or less.

When the driving scene of the vehicle is a specific driving scene, the driving skill evaluation ECU 10 stores predetermined data based on signals obtained from sensors and the like when the vehicle travels in a curve section as evaluation history data in the driving history database 10a. store in

On the other hand, the driving skill evaluation ECU 10 determines that the driving scene of the vehicle is not the specific driving scene when the acquired curve information (curvature radius R) and curve section average vehicle speed are not within the predetermined range AR1. Then, the driving skill evaluation ECU 10 discards the data based on the signals acquired from each sensor or the like when the vehicle travels in the curve section without storing it in the driving history database 10a.

(driving characteristic labeling)
Based on the evaluation history data, the driving skill evaluation ECU 10 evaluates the driving skill when the driving scene of the vehicle is a specific driving scene, and labels the evaluation history data with a driving characteristic label corresponding to the evaluation result. do. Then, the driving skill evaluation ECU 10 stores the labeled evaluation history data in the driving history database 10a. As a result, the driving characteristic label is stored in the driving history database 10a.

Specifically, the driving skill evaluation ECU 10 divides the driving characteristics (driving skill level) of the driver into two stages: driving characteristics A (for example, general level) and driving characteristics B (for example, high skill level). A label (label “0” or label “1”) corresponding to the determined result (driving characteristic A or driving characteristic B) is attached to evaluation history data for one specific scene unit.

In order to carry out the labeling, the driving skill evaluation ECU 10 assigns a first driving operation characteristic amount x1, a second driving characteristic amount x2, a third driving characteristic amount x3, and a fourth driving operation characteristic amount x1 to the evaluation history data. A feature amount x4 is calculated.

(First driving operation feature quantity x1)
The first driving operation feature amount x1 is a feature amount relating to the smoothness of the driving operation.

The driving skill evaluation ECU 10 uses FFT (Fast Fourier Transform) to decompose the time-series data of the steering angle periodically acquired by the steering angle sensor 14 into frequency domains. The driving skill evaluation ECU 10 calculates a value obtained by integrating a one-sided amplitude spectrum value of a predetermined frequency band, which is identified in advance by experiments or the like, in which rough driving is likely to appear, with respect to frequency components. Then, the driving skill evaluation ECU 10 uses the calculated value for x1.

Instead of the calculated value, the driving skill evaluation ECU 10 determines that the steering angular velocity is equal to or less than the steering angular velocity stop threshold (preliminarily determined by experiments, etc.) after the steering angular velocity exceeds the steering angular velocity operation threshold (previously specified by experiments, etc.). may be used as the first driving operation feature quantity x1. The number of times indicates the number of times that the driver's driving operation stops momentarily and becomes awkward. In this case, the driving skill evaluation ECU 10 has the advantage that the calculation load is low compared to the calculation of the frequency component.

Furthermore, the driving skill evaluation ECU 10 may use a value for evaluating the smoothness of the driving operation by the steering entropy method as the first driving operation characteristic quantity x1.

(Second driving operation feature amount x2, third driving operation feature amount x3, fourth driving operation feature amount)
The second driving operation feature amount x2, the third driving operation feature amount x3, and the fourth driving operation feature amount x4 represent the difference between the standard operation when the vehicle travels in the curve section and the driver's driving operation. is.

More specifically, as shown in Equation (1), the second driving operation feature amount x2 is the driver's turning operation time T2 and the standard turning operation time Tb2 when the vehicle travels in a curved section. is a ratio calculated by dividing the difference between the two by the reference operation time Tb2 of the cutting operation.

As shown in Equation (2), the third driving operation feature amount x3 is the difference between the driver's steering holding time T3 and the steering reference operation time Tb3 when the vehicle travels in a curve section. This is the ratio calculated by dividing by the standard operation time Tb3.

As shown in Equation (3), the fourth driving operation feature amount x4 is the difference between the driver's steering back operation time T4 and the standard steering operation time Tb4 when the vehicle travels in a curve section. This is the ratio calculated by dividing by the standard operation time Tb4 for the return operation.

The driving skill evaluation ECU 10 acquires the turning operation time T2, the steering holding time T3, and the steering operation time T4 based on the time-series data of the steering angle when the vehicle travels in the curve section.

The driving skill evaluation ECU 10 applies curve information (curvature radius R (m), curve distance D (m), transition curve length C (m)) when the vehicle travels in a curve section to equation (4). , the reference operation time Tb2 for the turning operation, the reference operation time Tb3 for holding the steering, and the reference operation time Tb4 for the steering return operation shown in FIG. For the coefficients a1 to a11, appropriate numerical values obtained in advance based on experiments or the like are applied.

（式（４）中、Ｒは曲率半径Ｒ（ｍ）であり、Ｄはカーブ距離Ｄ（ｍ）であり、Ｃは緩和曲線長（ｍ）であり、ａ１～ａ１１は係数である。）

(In formula (4), R is the radius of curvature R (m), D is the curve distance D (m), C is the transition curve length (m), and a1 to a11 are coefficients.)

Then, the driving skill evaluation ECU 10 calculates the acquired turning operation time T2, steering holding time T3, and steering back operation time T4, and the calculated reference turning operation time Tb2, steering holding reference operation time Tb3, and The reference operation time Tb4 of the switchback time is applied to formulas (1) to (3).

Thereby, the driving skill evaluation ECU 10 calculates the second driving characteristic amount x2, the third driving characteristic amount x3, and the fourth driving characteristic amount x4.

The driving skill evaluation ECU 10 applies the first to fourth driving operation feature amounts x1 to x4 calculated with respect to the evaluation history data to a classification model using logistic regression. As a result, the driving skill evaluation ECU 10 classifies the class to which the evaluation history data belongs into one of two classes (driving characteristic A and driving characteristic B) corresponding to two levels of driving skill levels. The driving skill evaluation ECU 10 labels the evaluation history data with a driving characteristic label (label "0" or label "1") corresponding to the classified class.

The classification model described above calculates a value (f (Ftotal)) for specifying the driving characteristic A and the driving characteristic B when the first to fourth driving operation feature amounts x1 to x4 are input, and the calculated value If (f(Ftotal)) is smaller than the determination threshold value Fth, it is determined (classified) as the driving characteristic A, and if the calculated value (f(Ftotal)) is equal to or greater than the determination threshold value Fth, it is determined as the driving characteristic B ( Classify).

Specifically, the driving skill evaluation ECU 10 applies the first to fourth driving operation feature quantities x1 to x4 to the logistic regression model, which is the equations (5) and (6). Thereby, the driving skill evaluation ECU 10 calculates a value (f(Ftotal)) for identifying (determining) the driving characteristic.

Ftotal=w1x1+w2x2+w3x3+w4x4 (5)

f(Ftotal)=1/(1+e- ^(Ftotal) ) (6)

When the calculated value f (Ftotal) is smaller than the determination threshold value Fth, the driving skill evaluation ECU 10 attaches a driving characteristic label (label "0") corresponding to the driving characteristic A to the evaluation history data, and stores the evaluation history data in the driving history database 10a. Store. On the other hand, when the calculated value f (Ftotal) is equal to or greater than the determination threshold value Fth, the evaluation history data is attached with a driving characteristic label (“1”) corresponding to the driving characteristic B and stored in the driving history database 10a.

Note that the parameters w1, w2, w3, and w4 in equation (5) and the determination threshold Fth are obtained by applying a large number of labeled evaluation history data collected in advance as follows to equations (5) and (6). is determined by performing statistical calculations.

A large number of historical data for evaluation of a specific driving scene of a driver identified as driving characteristic A is collected as historical data for evaluation labeled with driving characteristic A (label "0").
A large number of historical data for evaluation of specific driving scenes of the driver identified as driving characteristic B are collected as historical data for evaluation labeled with driving characteristic B (label "1").
・It should be noted that the certification of driving characteristics is performed by, for example, a dealer, a manufacturer, or the like.

In this example, the parameters w1 (fixed value), w2 (fixed value), w3 (fixed value) and w4 (fixed value) determined in this way, and the judgment threshold Fth (fixed value) are stored in the ROM. are stored and their fixed values are applied in equation (5). Note that the parameters w1, w2, w3 and w4 and the determination threshold value Fth may not be fixed values, and may be updated from time to time by values obtained from the outside via an external communication device (not shown) provided in the vehicle. may

Instead of a classification model using logistic regression, a classification model based on supervised machine learning (e.g., support vector machine or neural network) is created using historical data for evaluation with labels collected in advance as described above. Then, the evaluation history data may be labeled. Furthermore, in the above example, there are two classes into which the history data for evaluation is classified, but there may be three or more classes corresponding to driving skill levels divided into three or more stages.

(majority processing)
The driving skill evaluation ECU 10 stores the labeled evaluation history data in the driving history database 10a for each specific driving scene. Therefore, as the number of specific driving scenes increases, the number of collected (stored) driving characteristic labels increases. Every time the number of driving characteristic labels increases by a predetermined number, the driving skill evaluation ECU 10 determines the driving skill level corresponding to the largest number of driving characteristic labels by performing majority processing on the driving characteristic labels stored in the driving history database 10a. is determined as the final driving level of the driver.

<Specific operation>
The CPU (simply referred to as "CPU") of the driving skill evaluation ECU 10 executes the routine shown in the flow chart of FIG. 5 each time a predetermined period of time elapses.

Therefore, at a predetermined timing, the CPU starts processing from step 500 in FIG. 5, proceeds to step 505, and acquires data on the road on which the vehicle is traveling and data indicating the driving operation of the driver.

After that, the CPU proceeds to step 510 and determines whether or not the driving scene of the vehicle is the specific driving scene, as described above, based on the acquired data.

If the driving scene of the vehicle is not the specific driving scene, the CPU determines "No" in step 510, proceeds to step 595, and terminates this routine.

On the other hand, if the driving scene of the vehicle is the specific driving scene, the CPU determines "Yes" in step 510, performs steps 515 to 535 described below in order, and proceeds to step 595. Terminate this routine once.

Step 515: The CPU stores predetermined data among the data acquired from each sensor or the like when the vehicle travels in the curve section, in the driving history database 10a as history data for evaluation.
Step 520: As described above, the CPU calculates the first driving performance feature x1, the second driving performance x2, the third driving performance x3, and the fourth driving performance x4 with respect to the evaluation history data. do.
Step 525: As described above, the CPU calculates the calculated first driving operation feature amount x1, second driving operation characteristic amount x2, third driving operation characteristic amount x3, and fourth driving operation characteristic amount x4 according to the equation (5). and the value f (Ftotal) for specifying the driving characteristic is calculated by applying the equation (6). Furthermore, as described above, the CPU adds a driving characteristic label (label “0”) corresponding to the driving characteristic A and a driving characteristic B any of the driving characteristic labels (labeled "1") corresponding to . Then, the CPU stores the labeled evaluation history data in the driving history database 10a.
Step 530: The CPU, as described above, performs majority processing for the driving characteristic labels stored in the driving history database 10a.
Step 535: As described above, the CPU determines (updates) the driving characteristic corresponding to the largest number of driving characteristic labels as the driving skill level of the driver as a result of the majority processing.

<effect>
As described above, the present embodiment device evaluates the driving operation when the driving scene of the vehicle is the specific driving scene in the driving skill evaluation. Furthermore, the device of the present embodiment uses the driving operation feature quantities (x1 to x4) that easily show the difference in the driving skill of the driver to evaluate the driving skill. Furthermore, the embodiment device evaluates (labels) the driving skill for the evaluation history data obtained for each characteristic driving scene, and then, by majority vote processing, determines the driving skill level corresponding to the largest number of driving characteristic labels. is determined as the driving skill level of the driver. As described above, the device according to the embodiment can improve the accuracy of the driver's driving skill evaluation.

<Modification>
The present invention is not limited to the above embodiments, and various modifications can be adopted within the scope of the present invention.

For example, in the above-described embodiment, data indicating the motion of the vehicle (time-series data of the yaw rate and time-series data of the rate of change of the yaw rate) is acquired as the history data for evaluation, and the data representing the acquired motion of the vehicle has the following data. may be used instead of the first to fourth driving operation feature amounts x1 to x4 included in the data indicating the driver's driving operation.
- The first vehicle motion feature quantity is a feature quantity relating to the smoothness of changes in the yaw rate of the vehicle.
- The second vehicle driving feature amount is the increase time of the yaw rate change rate of the vehicle.
- The third vehicle driving characteristic amount is the time during which the yaw rate change rate of the vehicle is maintained.
- The fourth vehicle driving characteristic quantity is the decrease time of the yaw rate change rate of the vehicle.

Alternatively, for example, the processing of step 530 in the flowchart of FIG. , may be determined as the determination result. Also in this case, the accuracy of the driver's driving skill evaluation can be improved.

Reference Signs List 10 Driving skill evaluation ECU 11 Camera sensor 12 Accelerator pedal operation amount sensor 13 Brake pedal operation amount sensor 14 Steering angle sensor 15 Vehicle speed sensor 16 Acceleration sensor 17 Yaw rate sensor 18 … GPS receiver

Claims (1)

a detection unit for acquiring driving-related data including either data indicating the driving operation of the driver or data indicating the movement of the vehicle;
a storage unit including a driving history database for storing specific driving operation-related data out of the driving operation-related data acquired by the detection unit;
a driving skill evaluation unit that evaluates the driving skill of the driver by determining the driving skill level of the driver;
with
The driving skill evaluation unit
Acquiring the driving operation-related data when the vehicle travels in a curve section having a radius of curvature of a predetermined radius or less at a vehicle speed of a predetermined speed or more, as history data for evaluation;
Based on a predetermined feature amount of the history data for evaluation, classifying to which of a plurality of classes corresponding to the driving skill levels divided into a plurality of stages the class to which the history data for evaluation belongs belongs The class to which the acquired history data for evaluation belongs is determined using a predetermined classification model for determining the history data for evaluation, and the label corresponding to the determined class is attached to the history data for evaluation to determine the driving history. store in a database,
By performing majority vote processing for the labels stored in the driving history database, the driving skill level corresponding to the class corresponding to the largest number of the labels is determined as the driving skill level of the driver. configured to
In the driving skill evaluation device,
The detection unit is
configured to acquire at least steering angle data as the driving operation related data;
The storage unit
configured to store the time-series data of the driving operation-related data in the driving history database;
The driving skill evaluation unit
As the predetermined characteristic amount, a first driving characteristic amount (x1), a second driving characteristic amount (x2), a third driving characteristic amount, based on the time-series data of the steering angle as the history data for evaluation. Obtaining the amount (x3) and the fourth driving operation feature amount (x4),
here,
The first driving operation characteristic amount (x1) is obtained by decomposing the time-series data of the steering angle as the history data for evaluation into a frequency domain by a fast Fourier transform, and obtaining a one-sided amplitude spectrum value in a predetermined frequency band. is the value obtained by integrating the frequency component,
The second driving operation feature amount (x2) is the turning operation time (T2) of the driver when the vehicle travels in the curve section, the curvature radius (R) of the curve section, and the transition curve length (C ) is a ratio calculated by dividing the difference between the standard operation time (Tb2) of the cutting operation determined based on ) by the standard operation time (Tb2) of the cutting operation,
The third driving operation feature amount (x3) is based on the driver's steering holding time (T3) when the vehicle travels in the curve section and a curve distance (D) that is the distance of the curve section. It is a ratio calculated by dividing the difference between the standard operation time (Tb3) for holding the fixed steering by the standard operation time (Tb3) for holding the steering,
The fourth driving operation feature amount (x4) includes the steering operation time (T4) of the driver when the vehicle travels in the curve section, the curvature radius (R) of the curve section, and the transition curve length ( C) is a ratio calculated by dividing the difference between the reference operation time (Tb4) of the switchback operation determined based on C) by the reference operation time (Tb4) of the switchback operation,
Furthermore, as the predetermined classification model,
The first driving characteristic amount (x1), the second driving characteristic amount (x2), the third driving characteristic amount (x3), and the fourth driving characteristic amount (x4) are represented by w1, w2, and w3. and w4 is a given value,
F total = w1x1 + w2x2 + w3x3 + w4x4
The value Ftotal is obtained by applying the following formula, and the value Ftotal is
f(Ftotal)=1/(1+e- ^(Ftotal) )
Obtain a value (f (Ftotal)) for determining the driving characteristics by applying the formula
using a model that determines the class to which the acquired evaluation history data belongs based on a comparison between the value (f(Ftotal)) and a predetermined determination threshold (Fth);
configured as
Driving skill evaluation device.

Images