JP7778051B2 - Driving ability assessment system and driving ability assessment method - Google Patents

Description

The present invention relates to a driving ability assessment system and a driving ability assessment method for assessing the driving ability of a vehicle driver.

A known example of this type of device is one that measures a driver's safe driving ability (see, for example, Patent Document 1). This device intermittently applies a load to the driver through audio output to distract them, calculates steering entropy values that represent steering error under both a loaded and unloaded condition, and evaluates the driver's safe driving ability based on the difference between the error evaluation values calculated under the loaded and unloaded conditions.

By assessing the driving ability of elderly drivers and providing them with an opportunity to consider surrendering their driver's license or adopting driving assistance features as necessary, we can improve road safety and contribute to the development of sustainable transportation systems.

JP 2014-174848 A

However, the device described in Patent Document 1 above requires imposing a load on the driver in order to evaluate their safe driving ability, which interferes with driving.

A driving ability determination system according to one aspect of the present invention includes an information acquisition unit that acquires steering angle data indicating changes in the steering angle of the vehicle over time, along with vehicle size information; an evaluation value calculation unit that calculates a prediction error between a steering angle predicted at a second time point after the first time point based on the time-series steering angle data acquired up to a first time point and an actual steering angle acquired by the information acquisition unit at the second time point, and calculates an evaluation value indicating the magnitude of the driver's steering error based on the calculated prediction error; a driving ability determination unit that compares the evaluation value calculated by the evaluation value calculation unit with a predetermined judgment threshold and determines the driver's driving ability based on the comparison result; and a setting unit that sets at least one of a time interval from the first time point to the second time point and a judgment threshold based on the size information acquired by the information acquisition unit. The information acquisition unit acquires information on the wheelbase of the vehicle as size information. The setting unit sets the time interval to be shorter the shorter the wheelbase, and sets the judgment threshold to be larger the shorter the wheelbase, based on the wheelbase information acquired by the information acquisition unit.

Another aspect of the present invention relates to a driving ability assessment method, which includes the following steps, each executed by a computer: an information acquisition step of acquiring steering angle data indicating changes in the steering angle of the vehicle over time, along with vehicle size information; an evaluation value calculation step of calculating a prediction error between a steering angle predicted at a second time point after the first time point based on the time-series steering angle data acquired up to a first time point in the information acquisition step and an actual steering angle acquired at the second time point, and calculating an evaluation value indicating the magnitude of the driver's steering error based on the calculated prediction error; a driving ability assessment step of comparing the evaluation value calculated in the evaluation value calculation step with a predetermined judgment threshold and assessing the driver's driving ability based on the comparison result; and a setting step of setting at least one of a time interval from the first time point to the second time point and a judgment threshold based on the size information acquired in the information acquisition step. In the information acquisition step, information on the wheelbase of the vehicle is acquired as size information. In the setting step, the shorter the wheelbase, the shorter the time interval is set, and the larger the judgment threshold is set, based on the wheelbase information acquired in the information acquisition step.

This invention makes it possible to assess driving ability without interfering with driving.

FIG. 1 is a diagram for explaining a driving section and a driving load. 1 is a block diagram showing an example of a configuration of a main part of a driving ability determination system according to an embodiment of the present invention; FIG. 3 is a diagram for explaining fluctuations in the steering angle of a vehicle. FIG. 4 is a diagram for explaining a change in steering angle over time. 3 is a diagram for explaining setting of a calculation period of a prediction error by the setting unit in FIG. 2 . FIG. 10 is a diagram illustrating an example of a degree display of the degree of steering shake. 3 is a flowchart showing an example of processing executed by a calculation unit in FIG. 2 .

Embodiments of the present invention will now be described with reference to Figures 1 to 7. A driving ability determination system according to an embodiment of the present invention determines the driving ability of a vehicle driver. Generally, a driver's driving behavior is composed of three elements: cognition, judgment, and operation. Of these elements, abilities related to "cognitive function," which is a person's intellectual function related to cognition and judgment, are known to gradually decline with age. Declining cognitive function makes it difficult to drive a vehicle safely. The driving ability related to cognitive function can be determined based on driving data obtained when the driver drives a vehicle, allowing the driver and their family to understand the decline in cognitive function and support safe driving.

Figure 1 is a diagram explaining driving sections and driving load. As shown in Figure 1, the driving load imposed on a driver due to driving behavior changes depending on the driving section, such as road shape. For example, driving load increases when driving on S-curves or winding roads, or when parking in a parking space. That is, driving load increases on driving sections where the driver is required to make many steering maneuvers per unit of vehicle movement and the vehicle's driving trajectory has a complex shape. In this case, high driving skills are required, as steering is required in conjunction with accelerator and brake operations, and vehicle sense is also required. In such driving sections (high-load sections), the driver's driving skills have a significant impact on driving stability.

On the other hand, there is almost no driving load when driving on straight roads. In other words, there is almost no steering required from the driver per unit of vehicle movement, and the vehicle's driving trajectory is extremely simple, so there is almost no driving load. In such driving sections (no-load sections), the driver's driving skill has almost no effect on driving stability.

Driving on curved roads, changing lanes on multi-lane roads, and turning right or left at intersections are driving loads that are somewhere in between these two. Even in these driving sections (low-load sections), the driver's driving skill does not have much of an impact on driving stability.

However, even in no-load or low-load sections, for example, when driving on a single-lane highway (two-way traffic section) with no median strip, the driver needs to be aware of the conditions in their own lane and the oncoming lane. In this case, shifting their gaze between their own lane and the oncoming lane increases the driver's mental activity and increases the cognitive load (hereinafter referred to as "cognitive load"). Gaze shifts are also required in sections with multiple lanes with heavy traffic, sections with many signs and traffic signals, sections with many pedestrians such as in busy areas, sections with many blind spots such as curved roads with poor visibility or intersections, and sections where multiple roads intersect, thereby increasing the cognitive load.

Furthermore, when making a turn at an intersection to change the vehicle's direction of travel and cross the oncoming lane (a right turn in countries and regions where vehicles drive on the left, or a left turn in countries and regions where vehicles drive on the right; hereafter referred to simply as a "right turn"), the driver must grasp the situation of the oncoming lane ahead while also grasping the situation of the lane ahead after the right turn in order to recognize the vehicle's target trajectory. In this case, too, the driver's gaze must shift between the oncoming lane ahead and the lane ahead, increasing the cognitive load.

In driving sections where cognitive load is high (hereinafter referred to as "specific sections"), the state of the driver's cognitive function affects driving stability. By acquiring driving data for specific sections in a manner that allows them to be distinguished from other sections and evaluating driving stability based on that driving data, it is possible to determine the driver's driving ability related to their cognitive function. For example, it is possible to acquire time-series location information for the vehicle and, based on the acquired location information, identify driving data for specific sections that have been preset in map information.

However, generally, the shorter the vehicle's wheelbase (the length from the center of the front wheels to the center of the rear wheels), the more likely the vehicle body is to wobble, disrupting the driver's steering and making it more difficult to evaluate the driver's own driving stability based on driving data. Therefore, in this embodiment, the driving ability assessment system is configured as follows to appropriately assess driving ability related to cognitive function by changing the way driving data is handled depending on the vehicle's size (classification), such as the vehicle's wheelbase and overall length (the length of the vehicle in the front-to-rear direction). In the following description, an example will be described in which wheelbase is used as the vehicle size (classification). In the following description, a short/long vehicle wheelbase is synonymous with a short/long vehicle overall length and a small/large vehicle size (classification).

Figure 2 is a block diagram showing an example of the main components of a driving ability assessment system (hereinafter referred to as the system) 10. As shown in Figure 2, the system 10 includes a computer having a calculation unit 11 such as a CPU, a storage unit 12 such as ROM and RAM, and peripheral circuits. The calculation unit 11 functions as an information acquisition unit 13, a setting unit 14, a load assessment unit 15, an evaluation value calculation unit 16, a cognitive function assessment unit 17, and an information output unit 18. The storage unit 12 stores information such as programs executed by the calculation unit 11 and setting values. The system 10 may be configured as an on-board device installed in a vehicle, as a server device or the like provided external to the vehicle, or as a combination of an on-board device and an external server device or the like.

The information acquisition unit 13 acquires time-series vehicle driving data along with vehicle wheelbase information for each pre-registered driver. For example, it acquires wheelbase information for pre-registered vehicles that each driver drives on a daily basis, as well as driving data measured for that vehicle. The wheelbase information may be numerical wheelbase information or vehicle model information. The vehicle's wheelbase can be identified based on the vehicle model information. The vehicle wheelbase information acquired by the information acquisition unit 13 is stored in the storage unit 12 as user information for each pre-registered driver.

The driving data includes at least time-series steering angle data indicating changes in the steering angle θ of the steering wheel (steering wheel) by the driver, detected at a predetermined period (sampling period). The driving data also includes time-series position information of the vehicle. The driving data is transmitted from the vehicle to the system 10 at a predetermined period (communication period), for example, via a TCU (telematics control unit) installed in the vehicle. The information acquisition unit 13 acquires the driving data transmitted from pre-registered vehicles as time-series driving data for each pre-registered driver. The time-series driving data acquired by the information acquisition unit 13 is stored in the memory unit 12 together with user information as driving data for each pre-registered driver.

Based on the wheelbase information stored in the memory unit 12, the setting unit 14 sets at least one of the calculation period for the prediction error e(n) by the evaluation value calculation unit 16 (described later) and the judgment threshold used by the cognitive function evaluation unit 17 (described later).

The load determination unit 15 determines, for each unit of time, whether a predetermined cognitive load is acting on the driver based on the driving data stored in the memory unit 12. More specifically, the load determination unit 15 determines, based on the vehicle's time-series position information, whether the driving section for each unit of time is a no-load section or low-load section (no-load/low-load section) or a high-load section as shown in Figure 1. Furthermore, the load determination unit 15 determines whether the driving section for each unit of time is a specific section among the no-load/low-load sections that is preset in the map information.

The evaluation value calculation unit 16 calculates an α value representing the average magnitude of steering error for each individual driver and an Hp value representing the magnitude of steering error by the driver when cognitive load increases, based on the steering angle data stored in the memory unit 12. More specifically, the evaluation value calculation unit 16 calculates the α value based on steering angle data for a period determined by the load determination unit 15 to be traveling in a no-load/low-load section. The evaluation value calculation unit 16 also calculates the Hp value based on the calculated α value and steering angle data for a period determined by the load determination unit 15 to be traveling in a specific section.

Figure 3 is a diagram explaining fluctuations in the steering angle θ of a vehicle. When the vehicle is being driven stably, steering is smooth and the fluctuations in the steering angle θ are small. On the other hand, when the vehicle is being driven unstable, steering becomes shaky and the fluctuations in the steering angle θ become large.

More specifically, as shown in Figure 3, a predicted steering angle θp(n) at time n is calculated by second-order Taylor expansion centered on time (n-1) based on the actual steering angles θ(n-3), θ(n-2), and θ(n-1) at time points n-3, n-2, and n-1 immediately preceding time point n. Since the predicted steering angle θp(n) is an estimated value assuming smooth steering, if the actual steering is smooth, it will match the actual steering angle θ(n), but if the actual steering is unstable, it will deviate from the actual steering angle θ(n) depending on the degree of the unstable steering. The degree of such unstable steering can be expressed as a prediction error e(n) calculated by the following equation (i):
e(n)=θ(n)-θp(n)...(i)

Figure 4 is a diagram illustrating the change in steering angle θ over time. As shown in Figure 4, the amount of change per unit time in the vehicle's yaw angle and steering angle θ increases as the wheelbase becomes shorter, and decreases as the wheelbase becomes longer. Therefore, in order to calculate a significant prediction error e(n), the calculation period for prediction error e(n), i.e., the time interval from time n-1 to time n shown in Figure 3, needs to be sufficiently long the longer the wheelbase.

Figure 5 is a diagram illustrating the setting of the calculation cycle of the prediction error e(n) by the setting unit 14. The setting unit 14 calculates the optimal value of the calculation cycle of the prediction error e(n) based on the wheelbase information stored in the memory unit 12 and predetermined characteristics (for example, characteristics shown by the solid line in Figure 5). The optimal value is set to be shorter the shorter the wheelbase, and longer the longer the wheelbase.

When the optimal value calculated based on the wheelbase is within a predetermined range, the setting unit 14 sets the calculation period of the prediction error e(n) to the optimal value. More specifically, when the optimal value is equal to or greater than the sampling period of the steering angle θ and is within a predetermined appropriate range (e.g., approximately 100 ms to 300 ms) to ensure the calculation accuracy of the prediction error e(n), the calculation period of the prediction error e(n) is set to the optimal value. In this case, the calculation period of the prediction error e(n) is set to, for example, the value closest to the optimal value among integer multiples of the sampling period of the steering angle θ, and the predicted steering angle θp(n) and prediction error e(n) are calculated based on steering angle data that has been thinned out as necessary.

When the optimal value is greater than the upper limit of the predetermined range (the upper limit of the appropriate range in the example of Figure 5), the calculation period for the prediction error e(n) is set to the upper limit of the predetermined range. In this case, the calculation period for the prediction error e(n) is set to the maximum value within the appropriate range among values that are integer multiples of the sampling period of the steering angle θ, and the predicted steering angle θp(n) and prediction error e(n) are calculated based on steering angle data that has been thinned out as necessary.

When the optimal value is smaller than the lower limit of the predetermined range (the sampling period in the example of Figure 5), the calculation period for the prediction error e(n) is set to the lower limit of the predetermined range. In this case, the calculation period for the prediction error e(n) is set to the sampling period for the steering angle θ, and the predicted steering angle θp(n) and prediction error e(n) are calculated based on all steering angle data stored in the memory unit 12.

Figure 6 is a diagram illustrating an example of a frequency display of the degree of steering error, showing an example of a frequency display of the prediction error e(n). The evaluation value calculation unit 16 calculates the predicted steering angle θp(n) and prediction error e(n) at each time point n based on the steering angle data for the no-load and low-load sections, and calculates the 90th percentile value (α value) of the frequency distribution of the prediction error e(n) as shown by the solid line. The smoother the steering and the less steering error there is, the sharper the frequency distribution of the prediction error e(n) will be centered at "0°," where there is no steering error, and the smaller the α value will be. On the other hand, the more steering error there is, the broader the frequency distribution of the prediction error e(n) will be and the larger the α value will be.

By using steering angle data from no-load and low-load sections, excluding high-load sections where there is a lot of steering and driving skill has a large impact on steering error, it is possible to appropriately calculate the α value, which represents the magnitude of the driver's average steering error.

Furthermore, the evaluation value calculation unit 16 calculates the Hp value based on the calculated α value and the steering angle data for the specific section. More specifically, the evaluation value calculation unit 16 calculates the predicted steering angle θp(n) and prediction error e(n) for each time point n based on the steering angle data for the specific section, and divides the frequency distribution of the prediction error e(n) as shown by the dashed line into nine ranges P1 to P9 based on the α value. That is, based on the eight reference values -5α, -2.5α, -α, -0.5α, 0.5α, α, 2.5α, and 5α, the steering entropy value (Hp value) is divided into nine ranges: P1 (to -5α), P2 (-5α to -2.5α), P3 (-2.5α to -α), P4 (-α to -0.5α), P5 (-0.5α to 0.5α), P6 (0.5α to α), P7 (α to 2.5α), P8 (2.5α to 5α), and P9 (5α or higher). Then, based on the proportions p1 to p9 of each range P1 to P9, the steering entropy value (Hp value) is calculated using the following formula (ii):
Hp=-Σpi・log ₉ pi...(ii)

The Hp value represents the smoothness of steering, and decreases as steering tremor decreases and the frequency distribution of prediction error e(n) becomes sharper, while increasing as steering tremor increases and the frequency distribution of prediction error e(n) becomes broader. By using steering angle data from specific sections where there is a lot of eye movement and cognitive function is significantly affected by steering tremor, it is possible to appropriately calculate the Hp value, which represents the magnitude of the driver's steering tremor when cognitive load is higher than normal.

The cognitive function evaluation unit 17 evaluates the driver's cognitive function based on the Hp value calculated by the evaluation value calculation unit 16. More specifically, the cognitive function evaluation unit 17 compares the Hp value, which represents the magnitude of steering shake when cognitive load increases, with a predetermined judgment threshold, and determines, based on the comparison result, whether the driver's driving ability related to cognitive function meets or exceeds the standard for safely driving a vehicle. For example, if the Hp value calculated based on driving data from daily driving falls below the judgment threshold, it is evaluated that the driver's cognitive function has deteriorated and fallen below the standard for safely driving a vehicle.

The setting unit 14 sets the judgment threshold based on the wheelbase information stored in the memory unit 12. More specifically, when the optimal value of the calculation period of the prediction error e(n) calculated based on the wheelbase is within the predetermined range in FIG. 5 and the calculation period of the prediction error e(n) is set to the optimal value, the setting unit 14 sets the judgment threshold to a predetermined reference value (standard value). On the other hand, when the optimal value of the calculation period of the prediction error e(n) is outside the predetermined range in FIG. 5 and the calculation period of the prediction error e(n) is set to a value other than the optimal value, the setting unit 14 sets the judgment threshold to a value corrected from the reference value.

More specifically, when the optimal value for the calculation period of the prediction error e(n) is greater than the upper limit of the specified range in Figure 5 and the calculation period of the prediction error e(n) is set to the upper limit of the specified range (or a value close to that limit), the judgment threshold is set to a value corrected to be smaller than the reference value. In other words, when the wheelbase is long and the optimal value for the calculation period is greater than the upper limit of the appropriate range for ensuring the calculation accuracy of the prediction error e(n), it becomes difficult to observe a sufficient prediction error e(n). Therefore, by correcting the judgment threshold for the Hp value, which represents the magnitude of steering shake, to be smaller, the judgment criteria for the driver's cognitive function-related driving ability are made stricter.

Furthermore, when the optimal value for the calculation period of the prediction error e(n) is smaller than the lower limit of the specified range in Figure 5 and the calculation period of the prediction error e(n) is set to the lower limit of the specified range (or a value close to that limit), the judgment threshold is set to a value corrected to be larger than the reference value. In other words, when the wheelbase is short and the optimal value for the calculation period is smaller than the lower limit of the appropriate range for ensuring the calculation accuracy of the prediction error e(n) or the sampling period of the steering angle θ, a large prediction error e(n) is more likely to be observed. Therefore, by correcting the judgment threshold for the Hp value, which represents the magnitude of steering shake, to be larger, the judgment criteria for the driver's cognitive function-related driving ability are relaxed.

The information output unit 18 transmits the evaluation results from the cognitive function evaluation unit 17 to a user device such as the driver or a family member. For example, a notification can be sent to a pre-registered email address. In this case, the notification can prompt the driver or a family member to consider returning their driver's license or switching to a vehicle with enhanced driving assistance functions. Because objective information based on driving data is provided, it is easier for the driver to accept the current state of their cognitive function and consider appropriate measures early on.

Figure 7 is a flowchart showing an example of processing executed by the calculation unit 11 of the system 10. The processing shown in this flowchart is executed, for example, periodically. First, in step S1, wheelbase information and driving data stored in the memory unit 12 are read. Next, in step S2, an optimal value for the calculation cycle of the prediction error e(n) is calculated based on the wheelbase information read in step S1. Next, in step S3, it is determined whether the optimal value calculated in step S2 is within a predetermined range.

If the result in step S3 is affirmative, proceed to step S4, where the calculation period for the prediction error e(n) is set to the optimal value calculated in step S2, and the judgment threshold for the Hp value is set to a reference value. On the other hand, if the result in step S3 is negative, proceed to step S5, where the calculation period for the prediction error e(n) is set to a value within a predetermined range close to the optimal value calculated in step S2, and the judgment threshold for the Hp value is set to a value corrected from the reference value.

Next, in step S6, based on the driving data read in step S1, it is determined whether the driving section per unit time is a no-load/low-load section or a high-load section. Also, based on the driving data for the no-load/low-load section, it is determined whether the driving section per unit time is a specific section. Next, in step S7, based on the steering angle data for the period determined in step S6 to be driving in a no-load/low-load section, a predicted steering angle θp(n) and a predicted error e(n) are calculated at the calculation cycle set in step S4 or S5, and an α value is calculated. Also, based on the calculated α value and the steering angle data for the period determined in step S6 to be driving in a specific section, a predicted steering angle θp(n) and a predicted error e(n) are calculated at the calculation cycle set in step S4 or S5, and an Hp value is calculated. Next, in step S8, the Hp value calculated in step S7 is compared with the judgment threshold set in step S4 or S5, and the driver's cognitive function-related driving ability is evaluated based on the comparison result. Next, in step S9, the evaluation results from step S8 are sent to the pre-registered email address, and the process ends.

In this way, the α value and Hp value, which are indicators for assessing a driver's driving ability, can be calculated based solely on daily driving data, allowing driving ability to be assessed without interfering with driving (steps S1-S8). Furthermore, the driver's cognitive function is automatically assessed based solely on daily driving data, and the assessment results are notified to the driver and their family, reducing the burden on family members who live far away from the elderly driver (steps S1-S9). Furthermore, the calculation cycle for the prediction error e(n) and the assessment threshold for the Hp value are set taking into account the vehicle's wheelbase, allowing the driver's driving ability to be assessed appropriately regardless of the length of the wheelbase (steps S2-S5).

According to this embodiment, the following effects can be achieved.
(1) The system 10 includes an information acquisition unit 13 that acquires steering angle data indicating changes in the steering angle θ of the vehicle over time, along with information about the wheelbase of the vehicle; an evaluation value calculation unit 16 that calculates a prediction error e(n) between the steering angle θp(n) predicted at a time point n after time point n-1 and the actual steering angle θ(n) acquired at time point n, based on the time-series steering angle data acquired by the information acquisition unit 13 up to time point n-1, and calculates an evaluation value that represents the magnitude of the driver's steering shake based on the calculated prediction error e(n); a cognitive function evaluation unit 17 that compares the evaluation value calculated by the evaluation value calculation unit 16 with a predetermined judgment threshold and judges the driver's driving ability based on the comparison result; and a setting unit 14 that sets at least one of the time interval from time point n-1 to time point n (the calculation period for the prediction error e(n)) and the judgment threshold based on the wheelbase information acquired by the information acquisition unit 13 (FIG. 2).

In this way, by calculating an evaluation value that serves as an index for determining a driver's driving ability based on daily driving data, the driver's driving ability can be determined without interfering with driving. Furthermore, by setting the calculation cycle for the prediction error e(n) and the determination threshold for the Hp value taking into account the vehicle's wheelbase, the driver's driving ability can be appropriately determined regardless of the length of the wheelbase.

(2) The setting unit 14 sets a shorter calculation cycle for the prediction error e(n) the shorter the wheelbase, and sets a higher judgment threshold the shorter the wheelbase. In other words, for vehicles with a short wheelbase, setting a shorter calculation cycle for the prediction error e(n) of the steering angle θ reduces the impact of a large prediction error e(n) due to vehicle characteristics and improves the calculation accuracy of the evaluation value. In addition, by setting the judgment threshold for driving ability to a value corrected to be larger than the reference value, the impact of a large prediction error e(n) due to vehicle characteristics is taken into consideration and the judgment criteria for driving ability are relaxed.

(3) The setting unit 14 determines an optimal value for the calculation period of the prediction error e(n) based on the wheelbase information acquired by the information acquisition unit 13. If the determined optimal value is within a predetermined range, the calculation period of the prediction error e(n) is set to the optimal value and the judgment threshold is set to a reference value (steps S2 to S4 in FIG. 7). Furthermore, if the optimal value is greater than the upper limit of the predetermined range, the calculation period of the prediction error e(n) is set to the upper limit and the judgment threshold is set to a value smaller than the reference value. If the optimal value is smaller than the lower limit of the predetermined range, the calculation period of the prediction error e(n) is set to the lower limit and the judgment threshold is set to a value larger than the reference value (step S5). In this way, by prioritizing setting the calculation period of the prediction error e(n) to the optimal value whenever possible, the accuracy of the evaluation value calculation can be improved. By accurately calculating the evaluation value regardless of the length of the wheelbase, evaluation values calculated for vehicles of different vehicle classes can be appropriately compared.

(4) The specified range is equal to or greater than the sampling period of the time-series steering angle data acquired by the information acquisition unit 13 (Figure 5). That is, the calculation period for the prediction error e(n) is set to a value that is an integer multiple (1x or 2x or more) of the sampling period of the steering angle θ, and the predicted steering angle θp(n) and the prediction error e(n) are calculated based on all steering angle data or thinned-out steering angle data.

(5) The specified range is a range determined in advance to ensure the calculation accuracy of the prediction error e(n). In other words, the calculation period of the prediction error e(n) is set within a predetermined appropriate range (e.g., approximately 100 ms to 300 ms) to ensure the calculation accuracy of the prediction error e(n), thereby ensuring the calculation accuracy of the prediction error e(n).

(6) The system 10 further includes a load determination unit 15 that determines whether a predetermined cognitive load is acting on the driver based on the steering angle data acquired by the information acquisition unit 13 ( FIG. 2 ). The evaluation value calculation unit 16 calculates an α value representing the average magnitude of the driver's steering error based on the steering angle data acquired by the information acquisition unit 13, and calculates an Hp value representing the magnitude of the driver's steering error when the predetermined cognitive load is acting based on the calculated α value and the steering angle data acquired by the information acquisition unit 13 for a period during which the load determination unit 15 determines that the predetermined cognitive load is acting. The cognitive function evaluation unit 17 compares the Hp value calculated by the evaluation value calculation unit 16 with a determination threshold and determines the driver's driving ability related to cognitive function based on the comparison result. In this way, by calculating the Hp value representing the magnitude of the driver's steering error when the cognitive load is high based on driving data for a specific section where the cognitive load is high and comparing it with the determination threshold, it is possible to determine whether the driver's driving ability related to cognitive function is at or above a standard.

In the above embodiment, an example was described in which the α value was calculated based on steering angle data when traveling through a no-load/low-load section, and the Hp value was calculated based on steering angle data when traveling through a specific section, but the evaluation value calculation unit is not limited to this. The evaluation value calculation unit may be any unit that calculates an evaluation value that represents the magnitude of each driver's steering error based on the steering angle data for each driver.

The present invention has been described above as a driving ability assessment system, but it can also be used as a driving ability assessment method. That is, the driving ability assessment method includes an information acquisition step S1, executed by a computer, for acquiring steering angle data indicating changes in the vehicle's steering angle θ over time, along with information about the vehicle's wheelbase; an evaluation value calculation step S6, for calculating a prediction error e(n) between the steering angle θp(n) predicted at time n after time n-1 and the actual steering angle θ(n) acquired at time n, based on the time-series steering angle data acquired up to time n-1 in the information acquisition step S1, and calculating an evaluation value representing the magnitude of the driver's steering error based on the calculated prediction error e(n); a cognitive function assessment step S7, for comparing the evaluation value calculated in the evaluation value calculation step S6 with a predetermined judgment threshold and assessing the driver's cognitive driving ability based on the comparison result; and setting steps S4 and S5, for setting at least one of the time interval from time n-1 to time n (the calculation period for the prediction error e(n)) and the judgment threshold, based on the wheelbase information acquired in the information acquisition step S1 (FIG. 7).

The above description is merely an example, and the present invention is not limited to the above-described embodiments and variations, as long as the features of the present invention are not impaired. It is also possible to combine one or more of the above embodiments and variations in any desired manner, and it is also possible to combine variations together.

10. Driving ability assessment system (system), 11. Calculation unit, 12. Memory unit, 13. Information acquisition unit, 14. Setting unit, 15. Load assessment unit, 16. Evaluation value calculation unit, 17. Cognitive function assessment unit, 18. Information output unit

Claims (7)

an information acquisition unit that acquires steering angle data indicating a change in the steering angle of the vehicle over time together with information on the size of the vehicle;
an evaluation value calculation unit that calculates a prediction error between a steering angle predicted at a second time point after the first time point based on time-series steering angle data acquired up to a first time point by the information acquisition unit and an actual steering angle acquired at the second time point, and calculates an evaluation value that represents a magnitude of steering deviation by the driver based on the calculated prediction error; and
a driving ability determination unit that compares the evaluation value calculated by the evaluation value calculation unit with a predetermined determination threshold value and determines the driving ability of the driver based on the comparison result;
a setting unit that sets at least one of the time interval from the first time point to the second time point and the determination threshold value based on the size information acquired by the information acquisition unit ,
the information acquisition unit acquires information about a wheelbase of the vehicle as the size information,
A driving ability judgment system characterized in that the setting unit sets the time interval shorter the shorter the wheelbase, and sets the judgment threshold higher the shorter the wheelbase, based on the wheelbase information acquired by the information acquisition unit . 2. The driving ability determination system according to claim 1,
The information acquisition unit further acquires information about a length of the vehicle in a front-to-rear direction as the size information,
A driving ability determination system, characterized in that the setting unit sets at least one of the time interval and the determination threshold based on the length information acquired by the information acquisition unit. 2. The driving ability determination system according to claim 1,
The setting unit
determining an optimal value of the time interval based on the wheelbase information acquired by the information acquisition unit;
When the determined optimum value is within a predetermined range, the time interval is set to the optimum value and the determination threshold is set to a reference value;
When the optimum value is greater than an upper limit value of the predetermined range, the time interval is set to the upper limit value, and the determination threshold is set to a value smaller than the reference value;
A driving ability determination system characterized in that, when the optimal value is smaller than a lower limit value of the predetermined range, the time interval is set to the lower limit value and the determination threshold is set to a value greater than the reference value. 4. The driving ability determination system according to claim 3,
The driving ability determination system according to claim 1, wherein the predetermined range is equal to or greater than a sampling period of the time-series steering angle data acquired by the information acquisition unit. 5. The driving ability determination system according to claim 3,
A driving ability determination system, wherein the predetermined range is a range that is set in advance to ensure accuracy in calculating the prediction error. The driving ability determination system according to any one of claims 1 to 4,
a load determination unit that determines whether a predetermined load is acting on the driver based on the steering angle data acquired by the information acquisition unit,
the evaluation value calculation unit calculates a first evaluation value representing the magnitude of an average steering error of the driver based on the steering angle data acquired by the information acquisition unit, and calculates a second evaluation value representing the magnitude of the steering error of the driver when the predetermined load is applied based on the calculated first evaluation value and steering angle data for a period during which the load determination unit determines that the predetermined load is applied, among the steering angle data acquired by the information acquisition unit;
A driving ability determination system characterized in that the driving ability determination unit compares the second evaluation value calculated by the evaluation value calculation unit with the determination threshold value and evaluates the driver's driving ability based on the comparison result. Each of the methods is executed by a computer.
an information acquisition step of acquiring steering angle data indicating a change in steering angle of the vehicle over time together with information on the size of the vehicle;
an evaluation value calculation step of calculating a prediction error between a steering angle predicted at a second time point after the first time point based on the time-series steering angle data acquired up to the first time point in the information acquisition step and an actual steering angle acquired at the second time point, and calculating an evaluation value representing the magnitude of the driver's steering shake based on the calculated prediction error;
a driving ability determination step of comparing the evaluation value calculated in the evaluation value calculation step with a predetermined determination threshold value and determining the driving ability of the driver based on the comparison result;
a setting step of setting at least one of the time interval from the first time point to the second time point and the determination threshold value based on the size information acquired in the information acquisition step,
In the information obtaining step, information about a wheelbase of the vehicle is obtained as the size information,
A driving ability judgment method characterized in that in the setting step, the shorter the wheelbase, the shorter the time interval is set based on the wheelbase information acquired in the information acquisition step, and the larger the judgment threshold is set based on the wheelbase information acquired in the information acquisition step.