JP7114861B2 - Autonomous vehicle control system and control method - Google Patents

Description

The present invention relates to a control system and control method for an autonomous vehicle.

In recent years, research and development of self-driving vehicles have been actively carried out. Since self-driving vehicles are one type of vehicle that travels on roads, it is necessary to comply with laws and regulations such as the Road Traffic Act. Patent Document 1 is a technology related to vehicle legal compliance. This document discloses a device for determining that a vehicle's driving norm has been violated in an automated vehicle or a non-automated vehicle.

JP 2017-45327 A

On actual roads, there are various people, vehicles, etc., and they may take unexpected actions or actions that do not comply with laws and regulations. Therefore, compliance with laws and regulations may not be enough to make self-driving vehicles safer. SUMMARY OF THE INVENTION In view of the above background, the present invention provides a control system and control method for an automated driving vehicle with improved safety.

The present invention employs the following technical means in order to solve the above problems. The symbols in parentheses described in the claims and this section are an example showing the correspondence relationship with the specific means described in the embodiment described later as one aspect, and limit the technical scope of the present invention. not something to do.

The control system for an automated driving vehicle of the present invention includes a caution point database (14) generated by extracting caution points in driving scenes based on the data of laws (11) concerning road traffic or judicial precedents (12) concerning traffic accidents. and a driving scene calculation unit (23) for obtaining a driving scene of the automatically driving vehicle based on sensing data detected by a sensor (21) mounted on the automatically driving vehicle or a sensor (30) outside the automatically driving vehicle; A driving behavior determination unit that refers to the caution point database (14), extracts caution points corresponding to the driving scene obtained by the driving scene calculation unit (23), and determines a driving action based on the caution points. (24). In this way, by taking driving behavior in consideration of the precautions generated based on the law (11) and judicial precedent (12), it is possible to drive with a high degree of safety. In addition, driving can be controlled based on a unified standard instead of different standards for each driver.

The automatic driving vehicle control system of the present invention includes a caution database generation unit (13) that generates the caution database (14) based on the data of laws (11) regarding road traffic or judicial precedents (12) regarding traffic accidents. may be provided. With this configuration, when laws (11) and judicial precedents (12) are updated, it is possible to take driving behavior that reflects the contents thereof.

In the control system for an automated driving vehicle of the present invention, the driving scene calculator (23) may calculate a current or future driving scene based on the sensing data. With this configuration, the current or future driving scene can be predicted, and control can be performed to take driving behavior based on the points to be noted in the driving scene.

The control system for an automated driving vehicle of the present invention comprises a driving behavior model database (26) that stores driving behavior data corresponding to attention points, and the driving behavior determining unit (24) includes the driving behavior model database (26 ), the driving behavior may be determined by reading the driving behavior data corresponding to the attention point. As a result, it is possible to appropriately determine the driving behavior corresponding to the attention point.

In the automatic driving vehicle control system of the present invention, the caution database generation unit (13) judges that the vehicle should foresee risks among accidents that occur in the driving scene based on the court precedent (12). Reminders may be generated based on incidents reported. Accidents judged to be foreseeable by the vehicle according to court precedent (12) are the responsibility of the vehicle to avoid such accidents.

In the automatic driving vehicle control system of the present invention, the caution database (14) stores information on the magnitude of social loss when an accident occurs in association with the caution point, and the driving When a plurality of driving scenes are calculated by the scene calculating section (23), the driving action determining section (24) avoids an accident that may occur in each driving scene and causes a large social loss. Driving behavior may be determined. This configuration can reduce social loss due to accidents.

The control system for an automated driving vehicle of the present invention may include an information processing unit (22) that processes the sensing data into spatial description information that can be compared with the driving scene stored in the caution database (14). By processing the sensing data into the spatial description information in this way, it is possible to reduce the amount of calculation and the like when determining the driving behavior.

The control system for an automated driving vehicle of the present invention is a driving behavior model database (16 ), and a driving scene calculation unit (23) that obtains the driving scene of the automatically driving vehicle based on sensing data detected by a sensor (21) mounted on the automatically driving vehicle or a sensor (30) outside the automatically driving vehicle. , a driving behavior determination unit (28) that refers to the driving behavior model database (16), reads driving behavior data corresponding to the driving scene obtained by the driving scene calculation unit (23), and determines the driving behavior. ) and a driving behavior decision process data storage unit (29) for storing the driving scene data and the driving behavior data. By taking driving behavior based on the driving behavior model generated based on the law (11) and judicial precedent (12) in this way, it is possible to drive with a high degree of safety. In addition, driving can be controlled based on a unified standard instead of different standards for each driver. Furthermore, by storing the data of the driving scene and the driving behavior, the process of determining the driving behavior can be known later.

In the control system for an automatic driving vehicle of the present invention, the exercise action model database (16) stores data of caution points in the driving scene, and the driving action decision process data storage unit (29) stores the data in the driving scene. Data of caution points may be stored in association with each other. This makes it possible to know the process of determining what to pay attention to in each driving scene and determining the driving behavior.

The method for controlling an automatically driving vehicle of the present invention comprises a step (S10) of acquiring sensing data by a sensor (21) mounted on the automatically driving vehicle or a sensor (30) outside the automatically driving vehicle, and based on the sensing data , based on the step (S12) of obtaining the driving scene of the automatic driving vehicle, and the data of the law (11) concerning road traffic or the judicial precedent (12) concerning traffic accidents, attention generated by extracting caution points in the driving scene A point database (14) is referenced to extract attention points corresponding to the driving scene of the automatic driving vehicle, and a step (S14) of determining driving behavior based on the attention points. With this configuration, similar to the control system described above, it is possible to drive with a high level of safety based on unified criteria instead of different criteria for each driver.

The method for controlling an automatically driving vehicle of the present invention comprises a step (S30) of acquiring sensing data by a sensor (21) mounted on the automatically driving vehicle or a sensor (30) outside the automatically driving vehicle, and based on the sensing data , the step (S32) of determining the driving scene of the automated driving vehicle, and the driving behavior generated by determining the driving behavior in the driving scene based on the data of the law (11) on road traffic or the judicial precedent (12) on traffic accidents A step (S34) of reading driving action data corresponding to the driving scene of the automated vehicle with reference to the action model database (16) and determining the driving action; and a step of storing (S35). With this configuration, similar to the control system described above, it is possible to drive with a high level of safety based on unified criteria instead of different criteria for each driver. In addition, by storing data on driving scenes and driving behavior, the process of determining the driving behavior can be known later.

ADVANTAGE OF THE INVENTION According to this invention, the driving|operation with high safety|security which considered the caution generated based on laws and regulations (11) and judicial precedent (12) can be performed.

It is a figure which shows the structure of the control system of 1st Embodiment. It is a figure which shows the example of the data memorize|stored in caution DB of 1st Embodiment. (a) It is a figure which shows the example of a law. (b) It is a figure which shows the example of the caution generated from laws and regulations. (a) It is a figure which shows the example of a law and a judicial precedent. (b) is a diagram showing an example of notes generated from laws and judicial precedents. FIG. 4 is a diagram showing an example of data stored in a driving behavior model DB; FIG. It is a figure which shows operation|movement of the control system of 1st Embodiment. It is a figure which shows the example of the data memorize|stored in caution DB of 2nd Embodiment. It is a figure which shows operation|movement of the control system of 2nd Embodiment. It is a figure which shows the structure of the control system of 3rd Embodiment. FIG. 4 is a diagram showing an example of data stored in a driving behavior model DB; FIG. It is a figure which shows the example of the data memorize|stored in driving action determination process DB. It is a figure which shows operation|movement of the control system of 3rd Embodiment.

A control system for an automatic driving vehicle according to an embodiment of the present invention will be described below with reference to the drawings.
(First embodiment)
FIG. 1 is a diagram showing the configuration of a control system 1 for an automatically driving vehicle according to the first embodiment. The control system 1 for an automated driving vehicle according to the first embodiment is a system that generates data on points to be noted when driving based on laws and regulations, and controls driving behavior based on the data on points to be noted. be. The control system 1 includes a server 10 communicably connected to an automatically driving vehicle, and a calculation processing unit mounted on the automatically driving vehicle.

First, a general configuration of the automatic driving vehicle 20 will be described. The automatic driving vehicle 20 has an operation control unit 25 that controls automatic driving, various in-vehicle sensors 21 that detect the surrounding conditions of the vehicle, the operating state of the vehicle, etc., and a communication unit 27 that communicates with the external sensors 30. is doing. The in-vehicle sensor 21 is, for example, an acceleration sensor, a speed sensor, a steering angle sensor, a camera, a GPS, a navigation device, or the like. The vehicle exterior sensor 30 is, for example, a sensor mounted on a traffic light or a roadside device, a sensor of another vehicle, a security camera, or the like.

The operation control unit 25 is connected to vehicle equipment such as an engine, winkers, and lights. The operation control unit 25 performs operation control by controlling vehicle equipment of the automatically driven vehicle 20 using sensor data acquired by the in-vehicle sensor 21 and the outside sensor 30 . An autonomous vehicle travels in its lane according to the set route data so as not to collide with other vehicles or pedestrians, and observes traffic rules such as signals and traffic signs.

In order to further improve driving safety, the control system 1 of the present embodiment instructs the driving control unit 25 on driving behavior suitable for the driving scene. The configuration of the control system 1 will be described in detail below.

The server 10 includes a law database (hereinafter referred to as "law DB") 11 that stores laws related to road traffic, a precedent database (hereinafter referred to as "precedent DB" 12 that stores precedents related to traffic accidents, and driving precautions. and a precaution database (hereinafter referred to as "precaution DB") 14 that stores .

FIG. 2 is a diagram showing an example of data stored in the caution DB 14. As shown in FIG. The caution point DB 14 stores caution point data for each driving scene. For example, in a driving scene in which a car enters an intersection at a yellow light and turns right, and a pedestrian's traffic light is red, rear-end collision with a following vehicle is stored as a caution point.

A caution point database generation unit (hereinafter referred to as a “caution DB generation unit”) 13 has a function of generating a caution point DB 14 based on the data stored in the law DB 11 or the data stored in the precedent DB 12 .

FIG. 3(a) shows an example of a law, and FIG. 3(b) shows an example of caution points generated from the law. In this example, the caution point DB generation unit 13 determines that such a law is enacted because there are pedestrians who do not obey the red light, based on the legal data that "pedestrians must not cross the road at red lights." In view of the fact that the pedestrian traffic light has turned red, a caution point is generated for "a pedestrian who ignores the traffic light and crosses the street".

FIG. 4(a) shows an example of laws and regulations, and FIG. 4(b) shows an example of notes generated from the laws and regulations. Laws and regulations list places where you should drive slowly, such as road sections where road signs indicate that you should drive slowly, intersections where visibility is limited to the left and right, road turns, tops of uphill slopes, and steep downhill slopes. ing. Since it is a law, it is natural that these should be followed, but court precedent 1 has ruled that the duty to slow down is not exempted even on a road that is not very wide with a width of about 7.6 meters. That is, vehicles should foresee risks at intersections on roads with a width of 7.6 meters or less. Therefore, even in driving scenes that are not stipulated by laws and regulations, it is extracted that there is an obligation to slow down on roads with a width of about 7.6 meters or less.

In addition, case law 2 states that when entering an intersection where there is no visibility to the left or right, even when entering from a broad road that is generally considered to be a priority road, the duty to slow down is not exempted. Therefore, a wide road is not a priority road as defined in laws and regulations, and it is extracted as a point of caution that a driver should drive slowly when entering an intersection from a wide road. In this way, by also using judicial precedent data, it is possible to generate detailed precautions for driving scenes that cannot be covered by laws and regulations. Note that generation of caution points by the attention DB generation unit 13 can be performed by a known technique using artificial intelligence.

Next, the configuration of the calculation processing unit installed in the automatic driving vehicle 20 will be described. The information processing unit 22 generates spatial description information from sensing data detected by the vehicle interior sensor 21 and the vehicle exterior sensor 30 using a QSR (Qualitative Spatial Representative and Reasoning) technique. The spatial description information includes, for example, the lane in which the autonomous vehicle 20 is currently traveling, the presence or absence of an oncoming vehicle or a following vehicle, the distance to an intersection, the number of lanes, the width of the vehicle, the presence or absence of pedestrians, and the like. This is the information necessary for

A driving scene calculation unit 23 obtains a driving scene based on the data processed by the information processing unit 22 . In this embodiment, a future driving scene several seconds ahead is calculated. For example, if you are currently driving several tens of meters before an intersection and know from the navigation device's route data that you will turn right at the next intersection, the driving scene a few seconds ahead is to turn right at the intersection. is required. Also, if the traffic light at the intersection is currently green, for example, it is required that the traffic light is green or yellow when entering the intersection. When the data of signal change timing is received through road-to-vehicle communication, the state of the signal at the time of entering the intersection can be obtained with higher accuracy.

The driving behavior determining unit 24 has a function of reading out the caution points corresponding to the driving scene in which the automatic driving vehicle is currently placed from the caution point DB 14 and determining the driving behavior based on the read caution points. FIG. 5 is a diagram showing an example of data stored in the driving behavior model database (hereinafter referred to as "driving behavior model DB") 26. As shown in FIG. The driving behavior model DB 26 stores driving behavior data in association with caution points. For example, the driving behavior when caution should be exercised for a rear-end collision with a following vehicle is to move forward and maintain a distance from the following vehicle. In addition, the driving behavior to be careful of a collision with a pedestrian while walking on a crosswalk is to drive slowly at a speed at which the driver can stop immediately.

The driving behavior determining unit 24 determines the driving behavior by reading the driving behavior corresponding to the attention point from the driving behavior model DB 26 after finding the caution point in the driving scene. The driving behavior determination unit 24 transmits data on the determined driving behavior to the driving control unit 25 .

FIG. 6 is a diagram showing the operation of the control system 1 of this embodiment. The control system 1 of the autonomous driving vehicle acquires sensing data detected by the vehicle interior sensor 21 and the vehicle exterior sensor 30 (S10), processes the sensing data by the information processing unit 22, and generates space description information (S11). . Next, the driving scene calculator 23 calculates a future driving scene based on the current spatial description information (S12).

Subsequently, the driving action determination unit 24 receives the data stored in the caution DB 14 and determines whether or not there is a driving scene similar to the driving scene calculated by the driving scene calculation unit 23 (S13). . Here, if there is no similar driving scene (NO in S13), the control system 1 returns to step S10 for receiving sensing data.

If there is a similar driving scene (YES in S13), the driving action determining unit 24 reads caution points from the caution DB 14 associated with the similar driving scene. The driving behavior determination unit 24 determines a driving behavior based on the read caution points (S14). Specifically, from the driving behavior model DB 26, the driving behavior data corresponding to the points to be noted is read, and the driving behavior is determined. The driving action determination unit 24 transmits the determined driving action data to the driving control unit 25, and the driving control unit 25 performs driving control based on the driving action data (S15).

The configuration and operation of the control system 1 for an automatically driving vehicle according to the first embodiment have been described above. The control system 1 of the present embodiment reads caution points corresponding to the driving scene from the caution point DB 14 generated based on the law DB 11 and the judicial precedent DB 12, so it is possible to drive with high safety. Also, since the precautions are based on laws and judicial precedents, the grounds for driving behavior become clear.

In the present embodiment, the control system 1 has the caution DB 14 on the server 10 side, and the driving action determination unit 24 on the calculation processing unit side mounted on the automatic driving vehicle 20. However, the server 10 and the automatic driving vehicle 20 can be assigned functions as appropriate. For example, the automatic driving vehicle 20 may have the attention DB 14, and the automatic driving vehicle 20 may refer to the attention DB 14 to determine the driving behavior without communicating with the server 1. Conversely, by having the driving action determination unit 24 on the server 10 side and transmitting the data of the driving scene from the automatic driving vehicle 20 to the server, the server 10 side determines the driving action, and the automatic driving vehicle 20 side You may send.

(Second embodiment)
Next, the control system of the automatic driving vehicle of 2nd Embodiment is demonstrated. The basic configuration of the control system of the second embodiment is the same as the control system 1 of the first embodiment. The control system of the second embodiment is a system that takes appropriate driving behavior when a plurality of driving scenes are assumed.

FIG. 7 is a diagram showing an example of data stored in the caution DB 14. As shown in FIG. The caution DB 14 stores data on driving scenes and cautions, as well as loss index data. The data of the loss index is an index representing the magnitude of the loss when the accident described in the precautions occurs. In addition, the numbers described in FIG. 7 are given numbers for convenience of explanation.

FIG. 8 is a flow chart showing the operation of the control system for the automatic driving vehicle according to the second embodiment. FIG. 8 describes the flow of the portion corresponding to step S14 in the flow chart of the control system 1 in the first embodiment.

The driving scene calculator 23 obtains a plurality of driving scenes as future driving scenes. Since the future driving scene is calculated, it is assumed that it will not be fixed to one. For example, when entering an intersection and making a right turn, the indication of the traffic light when entering the intersection may be a green light or a yellow light. Moreover, when the traffic light is green, the signal for pedestrians may be green, or the green light may be blinking. FIG. 8 is a flow when several driving scenes are obtained by calculation in this way.

To explain using the example shown in FIG. 7, for example, when turning right at the next intersection, the driving scene is (R1) Entering the intersection at a yellow light and turning right. , (R2) Two driving scenes are calculated: (R2) entering an intersection with a green light, turning right, and the pedestrian traffic light blinking at that time. Also, for example, when turning left at the next intersection, the driving scene is (L1) entering the intersection with a yellow light and turning left, the pedestrian signal at that time is a red light, and (L2) entering the intersection with a green light. Then, two driving scenes are calculated: a left turn, and a pedestrian signal flashing at that time.

The driving action determination unit 24 selects one driving scene from among a plurality of assumed driving scenes (S20), and extracts caution points corresponding to the selected driving scene from the caution DB 14 (S21). It is determined whether or not this process has been performed for all possible driving scenes (S22). If it is determined that this process has not been performed for all driving scenes (NO in S22), step S20 for selecting a driving scene is performed. back to

If it is determined that all driving scenes have been processed (YES in S22), the driving action determination unit 24 determines driving actions corresponding to each driving scene (S23). Specifically, similarly to the first embodiment, the driving action data corresponding to the driving scene is read out from the driving action model DB 26 to determine the driving action.

The driving behavior determination unit 24 determines whether or not the driving behavior corresponding to each driving scene is the same (S24). The driving behavior is determined as the driving behavior that the automated driving vehicle 20 should take (S25). The example shown in FIG. 7 is an example of a left turn. In the driving scene (L1), the caution point is "an accident involving a two-wheeled vehicle on the rear side", and the driving action corresponding to the caution point is "slow down at a speed that allows immediate stopping" (see FIG. 5). In the driving scene (L2), the caution point is "collision with a pedestrian walking on a crosswalk", and the driving action corresponding to the caution point is "slow down at a speed that allows the driver to stop immediately" (Fig. 5). Therefore, although the assumed driving scenes are different, it is sufficient to take the same driving action, so the common driving action is determined as the driving action to be taken.

If the driving behavior differs depending on the driving scene (NO in S24), based on the loss index stored in the caution DB 14, the driving behavior corresponding to the point of caution with the largest loss index, that is, the driving behavior that avoids the maximum loss is selected. Select (S26). The example shown in FIG. 7 is an example of a right turn. In the driving scene (R1), the caution point is "rear-end collision with the following vehicle", and the driving action corresponding to the caution point is "go forward and keep a distance from the following vehicle" (see FIG. 5). ). In the driving scene (R2), the caution point is "collision with a pedestrian walking on a pedestrian crossing", and the driving action corresponding to the caution point is "slow down at a speed that allows the driver to stop immediately" (Fig. 5). In this way, the driving action to be taken differs depending on the driving scene. In this case, the loss indexes associated with the driving scene (R1) and the driving scene (R2) are compared, and the loss index of the driving scene (R2) is greater than the loss index of the driving scene (R1). (R2) Select a driving action corresponding to the attention point.

The driving action determination unit 24 transmits the determined driving action data to the driving control unit 25 and instructs the driving action.

The configuration and operation of the control system for the automatic driving vehicle according to the second embodiment have been described above. According to the control system of the second embodiment, when a plurality of driving scenes are assumed, it is possible to take driving behavior that reduces social loss due to accidents.

(Third Embodiment)
FIG. 9 is a diagram showing the configuration of a control system 3 for an automatic driving vehicle according to the third embodiment. In the automatic driving vehicle control system 3 of the third embodiment, the server 10 has a driving behavior model DB 16 instead of the caution DB 14 .

FIG. 10 is a diagram showing an example of data stored in the driving behavior model DB 16. As shown in FIG. The driving action model DB 16 stores data on driving actions to be taken in driving scenes. The driving behavior model DB 16 is generated by the driving behavior model DB generator 15 . The driving behavior model DB 16 extracts points of caution corresponding to driving scenes from the law DB 11 and case law DB 12, determines driving behaviors corresponding to the points of caution, and generates the driving behavior model DB 16. The driving behavior model DB 16 also stores data of precautions used when obtaining driving behavior corresponding to each driving scene. This data of points of caution is not directly necessary data when determining the driving behavior corresponding to the driving scene, but as will be described later, it is necessary to explain the driving behavior taken by the autonomous vehicle.

In the third embodiment, the driving behavior determination unit 28 reads the driving behavior corresponding to the driving scene obtained by the driving scene calculation unit 23 from the driving behavior model DB 16 and determines the driving behavior. A driving behavior determination process database (hereinafter referred to as “driving behavior determination process DB”) 29 is connected to the driving behavior determination unit 28 . The driving behavior determination unit 28 stores data indicating the process of determining the driving behavior in the driving behavior determination process DB 29 . At this time, the data of caution points corresponding to the driving scene is also stored in the driving behavior determination process DB 29 .

FIG. 11 is a diagram showing an example of data stored in the driving action determination process DB 29. As shown in FIG. The driving action decision process DB 29 stores the date and time, the driving scene, the data of the driving action, and the data of caution points. By referring to this data, it is possible to know what kind of driving action was taken in consideration of what kind of precautions in what kind of driving scene and when. This makes it possible to grasp the grounds for taking the driving action.

FIG. 12 is a flow chart showing the operation of the control system 3 for an automatically driven vehicle according to the third embodiment. The control system 3 of the automatic driving vehicle acquires sensing data detected by the vehicle interior sensor 21 and the vehicle exterior sensor 30 (S30), processes the sensing data by the information processing unit 22, and generates space description information (S31). . Next, the driving scene calculator 23 calculates a future driving scene based on the current spatial description information (S32).

Subsequently, the driving action determination unit 26 receives the data stored in the driving action DB 16 and determines whether or not there is a driving scene similar to the driving scene obtained by the driving scene calculation unit 23 (S33). . Here, if there is no similar driving scene, the control system 3 returns to step S30 for receiving sensing data.

If there is a similar driving scene (YES in S33), the driving behavior determination unit 28 reads the driving behavior from the driving behavior model DB 16 associated with the similar driving scene and determines the driving behavior (S34). Subsequently, the driving behavior determination unit 28 stores the driving scene for which the driving behavior was determined, the date and time when the driving scene was acquired, the driving behavior data, and the caution data in the driving behavior determination process DB 29 (S35). The driving action determination unit 24 transmits the determined driving action data to the driving control unit 25, and performs driving control based on the driving action data (S36).

The configuration and operation of the control system 3 for an automatic driving vehicle according to the third embodiment have been described above. The control system 3 of the third embodiment stores the data of the driving scene, the driving behavior, and the data of caution points, so that the process of determining the driving behavior can be known later. In the above-described embodiment, an example is given in which the data of caution points are also stored in the driving action determination process DB 29, but it is also conceivable that the data of caution points are not stored. This is because it is possible to obtain what kind of attention point the driving action was based on from the relationship between the driving scene and the driving action.

Although the control system and control method for an autonomous vehicle according to the present invention have been described in detail with reference to the embodiments, the control system and control method according to the present invention are not limited to the above-described embodiments. .

In the above-described embodiment, an example was given in which the driving scene calculator 23 obtains the future driving scene, but the current driving scene may be obtained. This allows the control system to determine driving behavior based on attention points in the current driving scene.

In the above-described embodiment, an example of having one caution point for one driving scene has been described, but a caution DB having a plurality of caution points for one driving scene may be used. An appropriate driving action may be determined from driving actions corresponding to a plurality of caution points for the driving scene calculated by the driving scene calculation unit. At this time, it is also possible to use a loss index as in the control system of the second embodiment.

Further, in the above-described embodiment, data on points of caution extracted from laws and regulations and judicial precedents is used to determine the driving behavior of an autonomously driven vehicle. It can also be applied to assist in driving. For example, while the driver is driving, information on points to be careful of according to the driving scene is provided by voice or the like. In addition, it is also possible to apply the data of caution points to the driving evaluation of whether or not the driver has been driving with attention to the caution points. For example, in each driving scene, the degree of safe driving of the driver is evaluated by comparing the driving behavior determined based on laws and judicial precedents with the driving behavior actually taken by the driver.

1, 3... Automatic driving vehicle control system, 10... Server, 11... Law DB,
12... Precedent DB, 13... Attention DB generation section, 14... Attention DB,
15...Driving behavior model DB generator, 16...Driving behavior model,
20 Automatic driving vehicle, 21 In-vehicle sensor, 22 Information processing unit,
23... driving scene calculation unit, 24... driving action determination unit, 25... driving control unit,
26...Driving action model DB, 27...Communication unit, 28...Driving action determination unit,
29...Driving action decision process DB, 30...Outside sensor

Claims (3)

a driving behavior model database (16) generated by determining driving behavior in driving scenes based on data from laws (11) on road traffic or judicial precedents (12) on traffic accidents;
Driving scene calculation for obtaining a plurality of possible future driving scenes of the autonomous vehicle based on sensing data detected by a sensor (21) mounted on the autonomous vehicle or a sensor (30) external to the autonomous vehicle. a part (23);
By referring to the driving behavior model database (16), driving behavior data corresponding to a plurality of future driving scenes obtained by the driving scene calculation unit (23) is read, and if each read driving behavior is the same. In some cases, a driving behavior determination unit (28) that determines a driving behavior common to each driving scene as a driving behavior to be taken ;
a driving action decision process data storage unit (29) for storing the driving scene data and the driving action data;
A control system for an automated driving vehicle. In the exercise action model database (16), data of cautions in the driving scene are stored,
2. The control system for an automatic driving vehicle according to claim 1, wherein said driving action decision process data storage unit (29) stores data of caution points in association with driving scenes. A step (S30) of acquiring sensing data by a sensor (21) mounted on an automatic driving vehicle or a sensor (30) outside the automatic driving vehicle;
obtaining a plurality of possible future driving scenes of the autonomous vehicle based on the sensing data (S32);
With reference to the driving behavior model database (16) generated by obtaining the driving behavior in the driving scene based on the data of the road traffic law (11) or the traffic accident precedent (12), the automated driving vehicle data of driving behaviors corresponding to a plurality of possible future driving scenes are read, and if the read driving behaviors are the same, the driving behavior common to each driving scene is determined as the driving behavior to be taken a step of determining a driving behavior (S34);
a step of storing the driving scene and the driving action data (S35);
A control method for an automated driving vehicle comprising:

Images