Autonomous Driving System

Abstract

To provide an automatic operation system in which a crewman can recognize whether a route planed based on a route planning model is a normal planning result of the route planning model.SOLUTION: An automatic operation system 100 plans a travel route through a route planning model 13a that is a machine learning model for planning a route of an autonomous travel of a vehicle. The automatic operation system 100 comprises: an automatic operation control part 13 which plans the travel route through the route planning model 13a based on detection result of an external sensor 2 and map information; a reference route calculation part 14 which calculates a reference route based on the detection result of the external sensor 2 or the map information; an influence degree calculation part 15 which calculates influence degree to be received by periphery of the vehicle through the autonomous travel of the vehicle along the planned route based on the comparison result between the planned route and the reference route; a threshold value calculation part 16 which calculates an influence degree threshold value that is a threshold value of the influence degree based on peripheral situation of the vehicle; and a notification control part 17 which notifies a crewman of the vehicle when the influence degree is the influence degree threshold value or more.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to automated driving systems. [Background technology]

[0002] Conventionally, a technology is known that predicts automatic driving that will cause shaking in the vehicle based on advance information obtained before executing automatic driving control, and notifies passengers of the vehicle behavior information that will cause the predicted shaking (for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2018-112850 Summary of the Invention [Problem to be solved by the invention]

[0004] In automated driving, where a route is planned using a route planning model, which is a machine learning model, a route may be planned that affects the surrounding area of ​​the vehicle. In this case, there is a possibility that the occupants may not be able to recognize whether such a route is a normal result of the route planning model. [Means for solving the problem]

[0005] One aspect of the present disclosure is an autonomous driving system that plans a route using a route planning model, which is a machine learning model for planning a route for autonomous driving of a vehicle, and includes: a route planning unit that plans a route using the route planning model based on detection results from the vehicle's onboard sensors and map information; a reference route calculation unit that calculates a reference route based on the detection results from the onboard sensors or the map information; an impact calculation unit that calculates the degree of impact on the surroundings of the vehicle due to the vehicle's autonomous driving along the planned route based on a comparison result between the planned route and the reference route; a threshold calculation unit that calculates an impact threshold, which is a threshold for the degree of impact, based on the surrounding conditions of the vehicle; and an alarm control unit that alerts the vehicle occupants when the degree of impact is equal to or greater than the impact threshold.

[0006] According to an automated driving system according to an aspect of the present disclosure, the degree to which a route resulting from a route planning model differs from a reference route is calculated as the degree of influence on the surroundings of the vehicle. For example, an influence threshold is calculated based on the validity of the difference between the route resulting from the route planning model and the reference route. In this manner, whether the route resulting from the route planning model conforms to the surrounding conditions of the vehicle can be estimated based on whether the influence is equal to or greater than the influence threshold. Therefore, even if a route planned by the route planning model affects the surroundings, by notifying the occupant that the route conforms to the surrounding conditions of the vehicle, the occupant can recognize whether such a route is a normal planned result of the route planning model.

[0007] In one embodiment, the notification control unit may notify other vehicles when there are other vehicles around the vehicle, so that the other vehicles around the vehicle can be notified that the route planned by the route planning model is in line with the surrounding conditions of the vehicle.

[0008] In one embodiment, when other vehicles around the vehicle are affected by the autonomous driving of the vehicle along the planned route, the notification control unit may notify only the affected other vehicles among the other vehicles. In this case, the other vehicles affected by the vehicle can be notified that the route planned as a result of the route planning model is in line with the surrounding conditions of the vehicle. [Effects of the Invention]

[0009] According to an autonomous driving system according to one aspect of the present disclosure, it becomes possible for an occupant to recognize whether a route planned by a route planning model is a normal planning result of the route planning model. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 is a block diagram illustrating an autonomous driving system according to an embodiment. [Figure 2] 10 is a flowchart illustrating an example of processing by an autonomous driving ECU. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.

[0012] FIG. 1 is a block diagram showing an autonomous driving system according to one embodiment. As shown in FIG. 1, the autonomous driving system 100 is installed in a vehicle such as a passenger car or a freight vehicle, and performs autonomous driving control of the vehicle. Autonomous driving control is vehicle control that causes the vehicle to autonomously travel along a planned route without the driver performing any driving operations. Note that some functions of the autonomous driving system 100 may be executed on a server that can communicate with the vehicle.

[0013] In the autonomous driving system 100, autonomous driving control is performed using a machine learning model. The autonomous driving system 100 plans a route using a route planning model, which is a machine learning model for planning a route for autonomous driving of a vehicle. The route planning model is a machine learning model that recognizes the external environment and generates a driving plan for autonomous driving control. In the autonomous driving system 100, the machine learning model may be used to generate instructions for autonomous driving control. The machine learning model will be described in detail below.

[0014] [Autonomous Driving System Configuration] The configuration of an autonomous driving system 100 according to this embodiment will be described below with reference to Fig. 1. As shown in Fig. 1, the autonomous driving system 100 includes an autonomous driving ECU (Electronic Control Unit) 10 that performs overall system management.

[0015] The autonomous driving ECU 10 is an electronic control unit having a central processing unit (CPU) and a storage unit. The storage unit is composed of, for example, read-only memory (ROM), random access memory (RAM), and electrically erasable programmable read-only memory (EEPROM). The autonomous driving ECU 10 realizes various functions by, for example, executing programs stored in the storage unit with the CPU. The autonomous driving ECU 10 may be composed of multiple electronic units.

[0016] The autonomous driving ECU 10 is connected to a GNSS receiver 1, an external sensor (on-board sensor) 2, an internal sensor (on-board sensor) 3, a map database 4, an actuator 5, an HMI [Human Machine Interface] 6, and an external notification unit 7.

[0017] The GNSS receiver 1 measures the position of the vehicle (for example, the latitude and longitude of the vehicle) by receiving signals from positioning satellites. The GNSS receiver 1 transmits the measured vehicle position information to the autonomous driving ECU 10.

[0018] The external sensor 2 is a detection device that detects the situation around the vehicle. The external sensor 2 includes at least one of a camera and a radar sensor. The camera is an imaging device that captures images of the situation outside the vehicle. The camera is installed, for example, behind the windshield of the vehicle and captures images in front of the vehicle. The camera transmits the captured images of the situation outside the vehicle to the autonomous driving ECU 10. The radar sensor is a detection device that detects objects around the vehicle using radio waves (for example, millimeter waves) or light. Radar sensors include, for example, millimeter wave radar or LiDAR (Light Detection and Ranging). The radar sensor transmits information about the detected objects to the autonomous driving ECU 10.

[0019] The internal sensors 3 are detection devices that detect the driving state of the vehicle. The internal sensors 3 include a vehicle speed sensor and a yaw rate sensor. The vehicle speed sensor is a detector that detects the speed of the vehicle. The vehicle speed sensor transmits the detected vehicle speed information (wheel speed information) to the autonomous driving ECU 10.

[0020] The yaw rate sensor is a detector that detects the yaw rate (rotational angular velocity) around the vertical axis of the center of gravity of the vehicle. For example, a gyro sensor can be used as the yaw rate sensor. The yaw rate sensor transmits the detected yaw rate information of the vehicle to the autonomous driving ECU 10.

[0021] The map database 4 is a database that stores map information. The map database 4 is formed, for example, in a storage device such as a hard disk drive (HDD) mounted on the vehicle. The map information includes road position information, road shape information (e.g., curves, types of straight sections, curvature of curves, etc.), intersection and branch point position information, and structure position information. The map database 4 may be formed in a server that can communicate with the vehicle.

[0022] The actuators 5 are devices used to control the vehicle. The actuators 5 include at least a drive actuator, a brake actuator, and a steering actuator. The drive actuator controls the amount of air supplied to the engine (throttle opening) in response to a control signal from the automatic driving ECU 10, thereby controlling the driving force of the vehicle. If the vehicle is a hybrid vehicle [HEV: Hybrid Electric Vehicle], in addition to the amount of air supplied to the engine, a control signal from the automatic driving ECU 10 is input to a motor serving as a power source to control the driving force. If the vehicle is an electric vehicle [BEV: Battery Electric Vehicle], a control signal from the automatic driving ECU 10 is input to a motor serving as a power source to control the driving force. In these cases, the motor serving as a power source constitutes the actuator 5.

[0023] The brake actuator controls the brake system in response to a control signal from the autonomous driving ECU 10, and controls the braking force applied to the vehicle wheels. The brake system may be, for example, a hydraulic brake system. The steering actuator controls the drive of an assist motor that controls the steering torque of the electric power steering system in response to a control signal from the autonomous driving ECU 10. In this way, the steering actuator controls the steering torque of the vehicle.

[0024] The HMI 6 is an interface for inputting and outputting information between the autonomous driving ECU 10 and the occupants. The HMI 6 includes, for example, a display, a speaker, and the like provided in the vehicle cabin. The HMI 6 outputs images from the display and sounds from the speaker in response to control signals from the autonomous driving ECU 10. The display may function as a touch panel. The display may be a center display, a navigation display, or a HUD (Head Up Display). The HUD presents information to the occupants by projecting images onto the windshield of the vehicle.

[0025] The external notification unit 7 is provided in the vehicle and outputs at least one of sound and light toward the outside of the vehicle. The external notification unit 7 has, for example, a sound output device that outputs sound outside the vehicle and a light output device that outputs light outside the vehicle. The light output device of the external notification unit 7 may be, for example, a lamp provided around the vehicle. The type of lamp is not particularly limited, and for example, a turn signal may be used. The light output device of the external notification unit 7 may also be a light output device such as an external display provided separately for notifying oncoming vehicles. The sound output device of the external notification unit 7 may also be, for example, a horn provided at the front of the vehicle. The sound output device of the external notification unit 7 may also be a sound output device such as an external speaker provided separately for notifying other vehicles around the vehicle.

[0026] Next, we will explain the functional configuration of the autonomous driving ECU 10. The autonomous driving ECU 10 has an external environment recognition unit 11, a driving state recognition unit 12, an autonomous driving control unit (route planning unit) 13, a reference path calculation unit 14, an influence degree calculation unit 15, a threshold calculation unit 16, and a notification control unit 17.

[0027] The external environment recognition unit 11 recognizes the external environment of the vehicle based on the detection results of the external sensor 2. The external environment includes the relative positions of surrounding objects with respect to the vehicle. The external environment may also include the relative speed and movement direction of surrounding objects with respect to the vehicle. The external environment may also include information on types of objects such as other vehicles, pedestrians, and bicycles. The external environment may also include information on lane markings recognized by known white line recognition or the like. The external environment recognition unit 11 may also recognize the external environment of the vehicle from the detection results of the external sensor 2 using a route planning model 13a.

[0028] The running state recognition unit 12 recognizes the running state of the vehicle based on the detection results of the internal sensor 3. The running state includes the vehicle speed, the vehicle acceleration, and the vehicle yaw rate. Specifically, the running state recognition unit 12 recognizes the vehicle speed based on the vehicle speed information from the vehicle speed sensor. The running state recognition unit 12 recognizes the vehicle acceleration based on the vehicle speed information from the acceleration sensor. The running state recognition unit 12 recognizes the direction of the vehicle based on the yaw rate information from the yaw rate sensor.

[0029] The automatic driving control unit 13 executes automatic driving control of the vehicle based on the external environment (surrounding environment) recognized by the external environment recognition unit 11 and the vehicle driving state recognized by the driving state recognition unit 12. The automatic driving control unit 13 may further use the vehicle position information measured by the GNSS receiving unit 1 and the map information in the map database 4 to perform automatic driving control.

[0030] The automatic driving control unit 13 performs automatic driving control of the vehicle using a route planning model 13a. The route planning model 13a is, for example, a neural network such as a convolutional neural network (CNN). The neural network can include multiple layers including multiple convolution layers and pooling layers. A deep learning network using deep learning can be used as the neural network. A recurrent neural network (RNN) can be used as the route planning model 13a.

[0031] The autonomous driving control unit 13 sets a target route for autonomous driving control based on, for example, a destination set by the occupant, vehicle position information, and map information. The target route is the route along which the vehicle will travel under autonomous driving control. The destination may be a point automatically proposed by the autonomous driving system 100. Alternatively, the target route may be set in advance in a well-known navigation system and acquired by the autonomous driving system 100 from the navigation system.

[0032] The autonomous driving control unit 13 generates a driving plan for autonomous driving control based on the target route, vehicle position information, and map information. The driving plan includes a vehicle speed plan and a vehicle route plan. The vehicle speed plan is data in which a target vehicle speed for autonomous driving control is associated with a set longitudinal position set at predetermined intervals (for example, 1 m) on the target route. Note that the set longitudinal position may be set based on the vehicle's driving time rather than distance. The route plan is data regarding a path on the target route. The route is a planned trajectory along which the vehicle under autonomous driving control is planned to travel on the target route. The route plan can be data on the steering angle or lateral position of the vehicle according to the set longitudinal position on the target route, for example.

[0033] The autonomous driving control unit 13 generates a driving plan using, for example, a route planning model 13a. The autonomous driving control unit 13 receives a target route, vehicle position information, and map information, and outputs the driving plan from the route planning model 13a. In this case, the route planning model 13a can be a neural network that has been trained to output a driving plan from the target route, vehicle position information, and map information.

[0034] The autonomous driving control unit 13 plans a route using the route planning model 13a based on the detection results of the external sensors 2 of the vehicle and map information. For example, the autonomous driving control unit 13 further inputs the current external environment and driving state into the route planning model 13a to re-output (modify) the driving plan. The route planned by the route planning model 13a here corresponds to the route of the driving plan modified in accordance with the external environment recognized based on the detection results of the external sensors 2. In this case, the route planning model 13a is a neural network that has been trained to output a driving plan using the target route, vehicle position information, map information, external environment, and driving state as inputs. The route planning model 13a may also receive the previously output driving plan as input.

[0035] The "modified route of the driving plan" refers to a modified route of the driving plan that affects the surroundings of the vehicle. For example, a "route that affects the surroundings of the vehicle" refers to a route of the driving plan that has been modified so that a new lane change, lane deviation, or deceleration of the vehicle is planned. Such modifications to the driving plan route may be based on the relative position and relative speed of the vehicle with respect to surrounding vehicles such as a preceding vehicle, or on the detection results of an external sensor 2, such as an obstacle on the road on which the vehicle is traveling. Such modifications to the driving plan route may also be based on map information, such as a reduction in the number of lanes on the road on which the vehicle is traveling, construction information on the road on which the vehicle is traveling, or the presence of a sharp curve ahead on the road on which the vehicle is traveling.

[0036] Such route corrections of the driving plan may include corrections in which the route planned by the route planning model 13a is in line with the vehicle's surrounding conditions, and corrections in which the route planned by the route planning model 13a is not in line with the vehicle's surrounding conditions. A correction in line with the vehicle's surrounding conditions means that there is no problem with the recognition processing and route planning processing by the route planning model 13a and the route is corrected in a reasonable manner in accordance with the vehicle's surrounding conditions. A correction that is not in line with the vehicle's surrounding conditions means that the route is not corrected in a reasonable manner in light of the vehicle's surrounding conditions, for example, due to some kind of calculation error occurring in the recognition processing or route planning processing by the route planning model 13a.

[0037] The automatic driving control unit 13 executes automatic driving control by transmitting a control signal to the actuator 5 based on the current vehicle position information and the driving plan, thereby changing the driving state of the vehicle.

[0038] The reference route calculation unit 14 calculates a reference route based on the detection results of the external sensor 2 or map information. The reference route means a route that serves as a reference for the route of the revised driving plan. For example, when the driving plan is revised in the autonomous driving control unit 13 based on the detection results of the external sensor 2, the reference route calculation unit 14 calculates the route before the revision as the reference route.

[0039] Specifically, assuming that calculation errors may occur in the recognition results of the lane markings by the route planning model 13a, the reference route calculation unit 14 may calculate, as the reference route, the extension direction of the lane markings included in the map information. Assuming that calculation errors may occur in the recognition results of the objects around the vehicle by the route planning model 13a, the reference route calculation unit 14 may calculate, as the reference route, a trajectory of a preceding vehicle in the lane in which the vehicle is traveling relative to the lane. Furthermore, assuming that calculation errors may occur in the vehicle speed plan planned by the route planning model 13a, the reference route calculation unit 14 may calculate, as the reference route, a trajectory corresponding to a range of vehicle speeds that the vehicle can take, based on the position of the vehicle in the lane in which the vehicle is traveling and the relative positions and relative speeds of other vehicles.

[0040] The influence calculation unit 15 calculates the degree of influence that the surroundings of the vehicle will have due to the autonomous driving of the vehicle along the planned route based on the comparison result between the planned route and the reference route. The influence is an index that represents the degree of influence of the "route that affects the surroundings of the vehicle." The greater the degree of influence, the greater the numerical value of the influence. The influence calculation unit 15 may calculate the influence to be greater the greater the distance (e.g., the distance in the lane width direction) that the lateral position of the planned route deviates from the lateral position of the reference route, for example, when the planned route bends significantly. The influence calculation unit 15 may calculate the influence to be greater the greater the distance (the error in the distance in the lane extension direction) that the longitudinal position on the planned route deviates from the longitudinal position on the reference route, for example, when the planned route includes deceleration. The influence calculation unit 15 may calculate the influence to be greater the closer the planned route is to the predicted trajectories of other vehicles around the vehicle, for example, when the planned route includes a lane change or the vehicle drifting into an adjacent lane. The influence degree calculation unit 15 may calculate the influence degree to be larger as the size of the range affected by the planned route becomes larger. The influence degree calculation unit 15 may calculate the influence degree to be larger as the number of other vehicles affected by the planned route becomes larger.

[0041] The threshold calculation unit 16 calculates an influence threshold based on the vehicle's surrounding conditions. The influence threshold is an influence threshold used to determine whether the route planned by the route planning model 13a is in accordance with the vehicle's surrounding conditions. When there are vehicle surrounding conditions that make it reasonable for the vehicle's route to be modified, the threshold calculation unit 16 calculates the influence threshold to be smaller than when there are no vehicle surrounding conditions that make it reasonable for the vehicle's route to be modified. When there are vehicle surrounding conditions that are reasonable enough to make it natural for the vehicle's route to be modified, the threshold calculation unit 16 may calculate the influence threshold to be smaller than when there are vehicle surrounding conditions that make it reasonable to modify the vehicle's route.

[0042] Specifically, when the threshold calculation unit 16 recognizes the presence of another vehicle parked at the left edge of the lane in which the vehicle is traveling so as to overhang the lane, the threshold calculation unit 16 may determine that there are circumstances around the vehicle that are reasonable enough to naturally cause the vehicle to decelerate and change its path to move toward the adjacent lane on the right, and may lower the influence threshold compared to before such recognition.When the threshold calculation unit 16 recognizes the presence of another vehicle that has begun to move toward the vehicle while traveling in the adjacent lane on the right, the threshold calculation unit 16 may determine that there are circumstances around the vehicle that are reasonable enough to cause the vehicle to change its path to move away from the other vehicle to the left or to decelerate and increase the distance between the other vehicle and the other vehicle, and may lower the influence threshold compared to before such recognition.When the threshold calculation unit 16 recognizes the illumination of the brake lights of a vehicle several vehicles ahead of the vehicle on a highway, the threshold calculation unit 16 may determine that there are circumstances around the vehicle that are reasonable enough to cause the vehicle to change its path to decelerate, and may lower the influence threshold compared to before such recognition. In addition, the threshold calculation unit 16 may adjust the impact threshold according to the trajectory type previously set by the occupant (for example, setting priority on safety, minimum travel time, etc.), and adjust the timing for determining whether the planned route is in line with the surrounding conditions of the vehicle according to the trajectory type.

[0043] When the degree of impact is equal to or greater than the impact threshold, the notification control unit 17 notifies the vehicle occupants. For example, the notification control unit 17 determines whether the degree of impact is equal to or greater than the impact threshold, and when it determines that the degree of impact is equal to or greater than the impact threshold, the notification control unit 17 notifies the vehicle occupants by, for example, text display or voice using the display or speaker of the HMI 6, that "the route is in line with the surrounding conditions of the vehicle."

[0044] The message "that the route is in line with the vehicle's surrounding conditions" may include the details of the route correction to the vehicle and the reasons, such as the vehicle's surrounding conditions, that made the route correction necessary. The details of the route correction to the vehicle may be displayed, for example, as an image in which the vehicle's trajectory before and after the correction is distinguished by solid and dashed lines, or may be explained by audio. The reason why the route correction to the vehicle is necessary may be displayed, for example, by highlighting on the HUD (for example, by enclosing in a rectangle) an object ahead of the vehicle that caused the route correction to occur. Other known methods may be used to convey the message "that the route is in line with the vehicle's surrounding conditions" as long as they are recognizable to the occupant.

[0045] The notification control unit 17 notifies other vehicles when there are other vehicles around the vehicle. For example, when other vehicles around the vehicle are affected by the autonomous driving of the vehicle along the planned route, the notification control unit 17 notifies only the affected other vehicles among the other vehicles. For example, the notification control unit 17 notifies only the affected other vehicles among the other vehicles by using the external notification unit 7 to output at least one of sound and light outside the vehicle. The sound output outside the vehicle may be a warning sound from a horn. The warning sound from a horn may be output continuously or intermittently. The light output outside the vehicle may be, for example, flashing of a turn signal toward other vehicles.

[0046] In addition, the notification control unit 17 may issue a notification to other vehicles around the vehicle when the other vehicles around the vehicle are not affected by the vehicle's autonomous driving along the planned route (when there are no other vehicles that will be affected).

[0047] Incidentally, the notification control unit 17 may notify the vehicle occupant when it is determined that the corresponding route needs to be corrected in accordance with a trajectory type preset by the occupant (for example, a setting of safety priority, minimum travel time, etc.). The notification control unit 17 may notify the vehicle occupant when it is determined that the route needs to be corrected by a steering speed or acceleration / deceleration of a certain amount or more.

[0048] The notification control unit 17 may issue a notification to other vehicles at the timing when a route adjustment including adjustment of a positional relationship shared between the vehicle and other vehicles during autonomous driving of the vehicle along a planned route is completed. Note that, if the planned route involves deceleration, and there is no following vehicle closer than a predetermined distance behind the vehicle, the notification control unit 17 may delay the timing of issuing a notification to following vehicles that are farther away than the predetermined distance.

[0049] [Autonomous driving system operation] Next, the operation of the automatic driving system 100 will be described with reference to the drawings. Fig. 2 is a flowchart showing an example of processing by the automatic driving ECU. The processing by the automatic driving ECU is executed, for example, while automatic driving control of the vehicle is being executed.

[0050] 2, in step S10, the autonomous driving ECU 10 of the autonomous driving system 100 plans a route using a route planning model using the autonomous driving control unit 13. The autonomous driving control unit 13 plans (modifies) a route using the route planning model 13a based on the detection results of the external sensors 2 of the vehicle and map information.

[0051] In step S11, the autonomous driving ECU 10 calculates a reference path using the reference path calculation unit 14. The reference path calculation unit 14 calculates the reference path based on the detection results of the external sensor 2 or map information.

[0052] In step S12, the autonomous driving ECU 10 calculates the degree of influence that the surroundings of the vehicle will have, using the influence calculation unit 15. Based on the comparison result between the planned route and the reference route, the influence calculation unit 15 calculates the degree of influence that the surroundings of the vehicle will have due to the autonomous driving of the vehicle along the planned route.

[0053] In step S13, the autonomous driving ECU 10 calculates an influence threshold value using the threshold value calculation unit 16. The threshold value calculation unit 16 calculates an influence threshold value that is a threshold value for the influence degree based on the surrounding conditions of the vehicle.

[0054] In step S14, the autonomous driving ECU 10 determines, via the notification control unit 17, whether or not the degree of impact is equal to or greater than the impact threshold. If the notification control unit 17 determines that the degree of impact is equal to or greater than the impact threshold (S14: YES), the autonomous driving ECU 10 proceeds to step S15. If the notification control unit 17 determines that the degree of impact is less than the impact threshold (S14: NO), the autonomous driving ECU 10 ends the current processing of FIG. 2. Thereafter, the autonomous driving ECU 10 repeats the processing from S10 again after a certain period of time has elapsed.

[0055] In step S15, the autonomous driving ECU 10 notifies the vehicle occupants via the notification control unit 17. The notification control unit 17 notifies the vehicle occupants that the planned route is in line with the surrounding conditions of the vehicle, for example, using a display or a speaker of the HMI 6.

[0056] In step S16, the autonomous driving ECU 10 determines, via the notification control unit 17, whether or not there is another vehicle around the vehicle. If the notification control unit 17 determines that there is another vehicle around the vehicle (S16: YES), the autonomous driving ECU 10 proceeds to step S17. If the notification control unit 17 determines that there is no other vehicle around the vehicle (S16: NO), the autonomous driving ECU 10 ends the current processing in FIG. 2. Thereafter, the autonomous driving ECU 10 repeats the processing from S10 again after a certain period of time has elapsed.

[0057] In step S17, the autonomous driving ECU 10 determines whether or not there is another vehicle that will be affected by the autonomous driving of the vehicle along the planned route using the notification control unit 17. If the notification control unit 17 determines that there is another vehicle that will be affected by the autonomous driving of the vehicle along the planned route (S17: YES), the autonomous driving ECU 10 proceeds to step S18.

[0058] In step S18, the autonomous driving ECU 10 causes the notification control unit 17 to notify only the other vehicles that will be affected. The notification control unit 17 notifies only the other vehicles that will be affected, for example, by using the external notification unit 7 to output at least one of sound and light toward the outside of the vehicle. Then, the current processing in FIG. 2 ends. Then, the autonomous driving ECU 10 repeats the processing from S10 again after a certain period of time has elapsed.

[0059] On the other hand, if the notification control unit 17 determines that there are no other vehicles that will be affected by the autonomous driving of the vehicle along the planned route (S17: NO), the autonomous driving ECU 10 proceeds to step S19.

[0060] In step S19, the autonomous driving ECU 10 causes the notification control unit 17 to notify other vehicles. The notification control unit 17 notifies other vehicles, for example, by using the external notification unit 7 to output at least one of sound and light toward the outside of the vehicle. Then, the current processing in FIG. 2 ends. Then, the autonomous driving ECU 10 repeats the processing from S10 again after a certain period of time has elapsed.

[0061] According to the above-described automated driving system 100, the degree to which the route resulting from the route planning model 13a differs from the reference route is calculated as the degree of influence on the surroundings of the vehicle. For example, an influence threshold is calculated based on the validity of the difference between the route resulting from the route planning model 13a and the reference route. In this manner, depending on whether the influence is equal to or greater than the influence threshold, it is possible to estimate whether the route resulting from the planning of the route planning model 13a is in line with the surrounding conditions of the vehicle. Therefore, even if the route planned by the route planning model 13a affects the surroundings, by notifying the occupant that the route is in line with the surrounding conditions of the vehicle, the occupant can recognize whether such a route is a normal planned result of the route planning model 13a.

[0062] In the autonomous driving system 100, when there is another vehicle around the vehicle, the notification control unit 17 notifies the other vehicle. This allows the other vehicles around the vehicle to be notified that the route planned by the route planning model 13a is in line with the surrounding conditions of the vehicle.

[0063] In the autonomous driving system 100, when other vehicles around the vehicle are affected by the autonomous driving of the vehicle along the planned route, the notification control unit 17 notifies only the affected other vehicles among the other vehicles. This makes it possible to notify other vehicles affected by the vehicle that the route planned by the route planning model 13a is in line with the surrounding conditions of the vehicle.

[0064] Although the embodiments of the present disclosure have been described above, the present disclosure is not limited to the above-described embodiments. The present disclosure can be implemented in various forms including the above-described embodiments and various modifications and improvements based on the knowledge of those skilled in the art.

[0065] In the above embodiment, for example, when the autonomous driving control unit 13 modifies the driving plan using the route planning model 13a based on the detection results of the external sensor 2, the reference route calculation unit 14 calculates the route before the modification as the reference route. However, this is not limited to this example. For example, the reference route calculation unit 14 may calculate, as the reference route, a route of the driving plan that has been planned without using the route planning model 13a, based on the detection results of the external sensor 2 or map information. "Without using the route planning model 13a" means that the route of the driving plan is planned using a known method that does not use a machine learning model. [Explanation of symbols]

[0066] 13...Automatic driving control unit (route planning unit), 13a...Route planning model, 14...Reference route calculation unit, 15...Influence calculation unit, 16...Threshold calculation unit, 17...Notification control unit, 100...Automatic driving system.

Claims

1. An autonomous driving system that plans a route using a route planning model that is a machine learning model for planning a route for autonomous driving of a vehicle, a route planning unit that plans the route using the route planning model based on detection results from an on-board sensor of the vehicle and map information; a reference route calculation unit that calculates a reference route based on the detection results of the on-board sensor or the map information; an influence calculation unit that calculates an influence on the surroundings of the vehicle due to the autonomous traveling of the vehicle along the planned route based on a comparison result between the planned route and the reference route; a threshold calculation unit that calculates an influence threshold that is a threshold of the influence degree based on a surrounding situation of the vehicle; An autonomous driving system comprising: an alert control unit that alerts an occupant of the vehicle when the impact level is equal to or greater than the impact level threshold.

2. The autonomous driving system according to claim 1 , wherein the notification control unit issues a notification to another vehicle when another vehicle is present around the vehicle.

3. 3. The autonomous driving system according to claim 2, wherein the notification control unit notifies only the other vehicles that are affected by the autonomous driving of the vehicle along the planned route when the other vehicles around the vehicle are affected.

Images