Vehicle control device and vehicle control method

Abstract

To solve such a problem that, in the case of traveling cooperatively with a mobile body existing in the surroundings, a sense of discomfort or fear for an occupant occurs due to closeness with a mobile body.SOLUTION: A vehicle control device comprises: an action estimation unit that estimates, on a road where an own vehicle and a mobile body other than the own vehicle exist, at least one or more action patterns of the mobile body; a route generation unit that defines, as a passable region, a region through which the own vehicle and the mobile body can simultaneously pass, and generates a route for the own vehicle to get to a target position within the passable region, with respect to the at least one or more action patterns of the mobile body estimated by the action estimation unit; a specific position setting unit that sets at least one or more specific positions on the route from a current position of the own vehicle to the target position; and an own vehicle behavior setting unit that sets speed of the own vehicle in a section from the current position of the own vehicle to each specific position on the basis of a predicted action probability indicating a likelihood that the mobile body will follow the action patterns estimated by the action estimation unit.SELECTED DRAWING: Figure 7

Description

[Technical Field]

[0001] The present invention relates to a vehicle control device and a vehicle control method. [Background technology]

[0002] In recent years, the development of autonomous driving technology for automobiles and other vehicles has progressed. However, when autonomous driving is performed, the road width on which the vehicle travels is not always appropriate. For example, on narrow roads where the road width is narrow and it is difficult for the vehicle and oncoming vehicles to pass each other, control different from that used during normal driving is required to move the vehicle to a place where they can pass each other.

[0003] For example, Patent Document 1 discloses an information processing device that controls the starting, stopping, direction of a vehicle on a road where there is no turning-off space between the vehicle and an oncoming vehicle, so that the vehicle returns to the entrance of the road where the two vehicles can pass each other. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2018-151177 Summary of the Invention [Problem to be solved by the invention]

[0005] Incidentally, when performing automated passing driving on narrow roads in urban areas or automated parking assistance in parking lots, the vehicle predicts the positions of surrounding vehicles, such as oncoming vehicles, and passes them at those positions. Here, there is a possibility that the oncoming vehicle may change its driving intention, or that an error may occur in the vehicle's prediction of the oncoming vehicle's behavior. In such cases, it becomes necessary to change the driving route during passing driving or automated parking assistance, which reduces convenience.

[0006] In view of these points, the present invention aims to provide a vehicle control device and a vehicle control method that can alleviate the discomfort felt by occupants and improve driving efficiency based on the results of predicting the behavior of a moving body when driving cooperatively with a moving body in the vicinity. [Means for solving the problem]

[0007] In order to solve the above problems, for example, the configurations described in the claims are adopted. The present application includes multiple means for solving the above-mentioned problems, and one example thereof is a vehicle control device comprising: a behavior estimation unit that estimates at least one behavior pattern of a moving body on a road on which the host vehicle and other moving bodies exist; a route generation unit that defines an area through which the host vehicle and the moving body can pass simultaneously as a passable area and generates a route for the host vehicle to a target position within the passable area for the at least one behavior pattern of the moving body estimated by the behavior estimation unit; a specific position setting unit that sets at least one specific position on the route from the current position of the host vehicle to the target position; and a host vehicle behavior setting unit that sets the speed of the host vehicle in the section from the current position of the host vehicle to the specific position based on a behavior prediction probability that indicates the possibility that the moving body will perform the behavior pattern estimated by the behavior estimation unit. [Effects of the Invention]

[0008] According to the present invention, even when it becomes necessary to change the driving route while passing another vehicle due to a change in the intention of an oncoming vehicle in the vicinity or an error in the prediction of the oncoming vehicle's behavior predicted by the vehicle itself, it is possible to plan a trajectory that alleviates the discomfort felt by the vehicle's occupants and improves driving efficiency. Problems, configurations, and effects other than those described above will become apparent from the following description of the embodiments. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a block diagram showing an example of the configuration of a driving system and sensors of a vehicle equipped with a vehicle control device according to a first embodiment of the present invention. [Figure 2]1 is a block diagram showing an example of the configuration of a vehicle control device according to a first embodiment of the present invention. [Figure 3] 1 is a block diagram showing an example of the configuration of a risk map generation unit of a vehicle control device according to a first embodiment of the present invention. FIG. [Figure 4] 5 is a flowchart showing an example of a delay risk map generation process performed by the vehicle control device according to the first embodiment of the present invention. [Figure 5] 3A and 3B are diagrams showing an example of a cooperative action plan and a congestion risk map generated in the vehicle control device according to the first embodiment of the present invention. [Figure 6] 1 is a block diagram showing an example of the configuration of a vehicle operation planning unit of a vehicle control device according to a first embodiment of the present invention. [Figure 7] 1 is a block diagram showing an example of the configuration of a cooperative behavior trajectory generation unit of a vehicle control device according to a first embodiment of the present invention. [Figure 8] 5 is a flowchart illustrating an example of a process for generating a turning-off position and attitude by the vehicle control device according to the first embodiment of the present invention. [Figure 9] 5A to 5C are diagrams illustrating an example of a process for generating candidate target turning-off position and attitude by a vehicle control device according to a first embodiment of the present invention. [Figure 10] 5 is a flowchart showing an example of a cooperative behavior trajectory generation process performed by the vehicle control device according to the first embodiment of the present invention. [Figure 11] 3 is a diagram showing an example of an arrival position candidate by a vehicle control device according to a first embodiment of the present invention. FIG. [Figure 12] FIG. 2 is a diagram showing an example of a cooperative behavior trajectory (passing trajectory) by the vehicle control device according to the first embodiment of the present invention. [Figure 13] FIG. 2 is a diagram showing an example of a retention risk map and a planned cooperative behavior trajectory (passing trajectory) by the vehicle control device according to the first embodiment of the present invention. [Figure 14] FIG. 10 is a diagram showing an example in which a passenger in the host vehicle feels uneasy or uncomfortable when traveling along a cooperative behavior trajectory according to the first embodiment of the present invention. [Figure 15]FIG. 1 is a block diagram showing an example of the configuration of a cooperative behavior speed profile calculation unit that takes into account a case where the behavioral intention of an oncoming vehicle has not been determined by a vehicle control device according to a first embodiment of the present invention. [Figure 16] FIG. 1 is a diagram showing speed branch points in consideration of a case where the behavioral intention of an oncoming vehicle has not been determined by a vehicle control device according to a first embodiment of the present invention. [Figure 17] FIG. 4 is a diagram showing an example of a speed profile in consideration of a case where the behavioral intention of an oncoming vehicle has not been determined by a vehicle control device according to a first embodiment of the present invention. [Figure 18] 1 is a block diagram showing an example of the configuration of a driving mode management unit of a vehicle control device according to a first embodiment of the present invention. [Figure 19] FIG. 10 is a block diagram showing an example of the configuration of a cooperative behavior speed profile calculation unit that takes into account a case where the behavioral intention of an oncoming vehicle has not been determined by a vehicle control device according to a second embodiment of the present invention. [Figure 20] FIG. 10 is a diagram showing an example of a speed branch point in consideration of a case where the behavioral intention of an oncoming vehicle has not been determined by a vehicle control device according to a second embodiment of the present invention. [Figure 21] FIG. 10 is a diagram showing an example of a speed profile in consideration of a case where the behavioral intention of an oncoming vehicle has not been determined by a vehicle control device according to a second embodiment of the present invention. [Figure 22] FIG. 10 is a diagram showing an example of a standby position candidate according to a first modified example of the embodiment of the present invention. [Figure 23] FIG. 10 is a diagram showing an example of a standby position candidate according to a second modification of the embodiment of the present invention. [Figure 24] FIG. 10 is a diagram showing an example of a route branch point according to a third modified example of the embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, examples of modes for carrying out the present invention (hereinafter referred to as "embodiments") will be described with reference to the accompanying drawings.

[0011] In this specification and the accompanying drawings, identical or similar components are denoted by the same reference numerals, and redundant explanations may be omitted or only explanations focusing on the differences may be given. The number of each component may be singular or plural unless otherwise specified.

[0012] First Embodiment A vehicle control device and a vehicle control method according to a first embodiment of the present invention will be described below with reference to FIGS.

[0013] [Vehicle configuration] First, the overall configuration of a vehicle equipped with a vehicle control device according to a first embodiment of the present invention will be described with reference to FIG. 1 is a diagram showing an example of the configuration of a driving system and sensors of a vehicle 100 equipped with a vehicle control device 1 according to the first embodiment. The vehicle 100 performs automatic driving under the control of the vehicle control device 1.

[0014] As shown in FIG. 1, a vehicle 100 has a left front wheel 101FL, a right front wheel 101FR, a left rear wheel 101RL, and a right rear wheel 101RR. The vehicle 100 is equipped with a forward recognition sensor 2, a left side recognition sensor 3, a right side recognition sensor 4, and a rearward recognition sensor 5 as sensors for recognizing the outside world, and is able to detect the relative distance and relative speed between the vehicle and surrounding vehicles. For example, the vehicle 100 can use a camera as a sensor for recognizing the outside world. Information (detection signals) from these sensors 2, 3, 4, and 5 is supplied to the vehicle control device 1.

[0015] The vehicle control device 1 calculates command values ​​for a steering control mechanism 10, a brake control mechanism 13, and a throttle control mechanism 20 to control the traveling direction of the vehicle 100 based on information from sensors 2, 3, 4, and 5. The vehicle 100 is equipped with a steering control device 8 that controls the steering control mechanism 10 based on command values ​​from the vehicle control device 1, and a braking control device 15 that controls the brake control mechanism 13 and adjusts the brake force distribution to each wheel based on command values ​​from the vehicle control device 1. The vehicle 100 is further equipped with an acceleration control device 19 that controls the throttle control mechanism 20 and adjusts the torque output of the engine based on command values ​​from the vehicle control device 1, and a display device 24 that displays a travel plan for the vehicle 100 and predicted behavior of nearby moving objects, etc.

[0016] The vehicle 100 also includes a communication device 23 that performs road-to-vehicle or vehicle-to-vehicle communication. Note that the sensor configuration shown in Fig. 1 is an example and is not limited to the example in Fig. 1. The types of sensors may also be various types of sensors such as ultrasonic sensors, stereo cameras, infrared cameras, radar, and LiDAR, or a combination of these sensors.

[0017] The vehicle control device 1 is configured as an arithmetic processing device including, for example, a CPU, a memory, etc., as will be described later with reference to Fig. 2. The vehicle control device 1 stores a program for performing vehicle driving control processing, and the vehicle control device 1 generates a driving plan by executing the program. The vehicle control device 1 calculates command values ​​for each control mechanism (hereinafter also referred to as "actuators") 10, 13, 20 for controlling vehicle driving in accordance with the generated driving plan. The control devices 8, 15, 19 for each actuator 10, 13, 20 receive the command values ​​of the vehicle control device 1 via communication and control each actuator based on the command values.

[0018] The configuration for operating the brake shown in FIG. 1 will be described. When the driver is driving, the pedal force generated by the driver stepping on the brake pedal 12 is boosted by a brake booster (not shown), and a corresponding hydraulic pressure is generated by a master cylinder (not shown). The generated hydraulic pressure is supplied via a brake control mechanism 13 to wheel cylinders 16FL, 16FR, 16RL, and 16RR arranged on each of the wheels 101FL to 101RR.

[0019] Wheel cylinders 16FL to 16RR are each composed of a cylinder, a piston, a pad, etc. The piston is propelled by hydraulic fluid supplied from master cylinder 9, and the pad connected to the piston is pressed against the disc rotor. The disc rotor rotates together with the wheels constituting each of the wheels 101FL to 101RR. Therefore, the brake torque acting on the disc rotor becomes a braking force acting between the wheel and the road surface. With the above configuration, braking force can be generated on each wheel in response to the driver's brake pedal operation.

[0020] The braking control device 15 is configured as an arithmetic processing device including, for example, a CPU and a memory, similar to the vehicle control device 1. The braking control device 15 receives inputs from a combined sensor 14 capable of detecting longitudinal acceleration, lateral acceleration, and yaw rate, wheel speed sensors 11FL to 11RR installed on each wheel, a braking force command from the braking control device 15, and a sensor signal from a steering wheel angle detection device 21 via a steering control device 8, which will be described later. The output of the braking control device 15 is supplied to a brake control mechanism 13 having a pump and a control valve, and the braking control device 15 can generate any braking force on each wheel independently of the driver's brake pedal operation.

[0021] The braking control device 15 estimates vehicle spin, drift-out, and wheel lock based on the input of sensor signals, and generates braking forces on the corresponding wheels to suppress these, thereby improving the driver's driving stability. Furthermore, the braking control device 15 can generate any braking force on the vehicle 100 upon receiving a brake command from the vehicle control device 1. This allows the braking control device 15 to automatically perform braking in autonomous driving where no driver operation is required. However, the braking configuration is not limited to this configuration, and other actuators such as a brake-by-wire may also be used.

[0022] Next, the configuration for performing the steering operation will be described. While the driver is driving the vehicle 100, the steering torque detection device 7 detects the steering torque input by the driver via the steering wheel 6, and the steering wheel angle detection device 21 detects the steering wheel angle. Based on this information, the steering control device 8 controls the motor to generate an assist torque. Note that, like the vehicle control device 1, the steering control device 8 is also configured as an arithmetic processing device including, for example, a CPU, a memory, etc.

[0023] The resultant force of the driver's steering torque and the assist torque from the motor moves the steering control mechanism 10, changing the direction of the front wheels (turning the front wheels). Meanwhile, depending on the turning angle of the front wheels, a reaction force from the road surface is transmitted to the steering control mechanism 10, and is then transmitted to the driver as a road reaction force.

[0024] The steering control device 8 generates torque using a motor and controls the steering control mechanism 10, independently of the driver's steering operation. Meanwhile, the vehicle control device 1 controls the front wheels to a desired turning angle by sending a steering force command to the steering control device 8. In this way, the vehicle control device 1 automatically performs steering in autonomous driving where no driver operation is required. However, this embodiment is not limited to the steering control device 8 shown in FIG. 1, and a configuration may be adopted in which steering operation is performed using another actuator such as a steer-by-wire.

[0025] Next, the configuration of the accelerator will be described. The amount of depression of the accelerator pedal 17 by the driver is detected by a stroke sensor 18 and input to an acceleration control device 19. Like the vehicle control device 1, this acceleration control device 19 is also configured as an arithmetic processing device including, for example, a CPU, a memory, etc.

[0026] The acceleration control device 19 controls the engine by adjusting the throttle opening according to the depression amount of the accelerator pedal 17. In this way, the acceleration control device 19 can accelerate the vehicle 100 according to the accelerator pedal operation by the driver. Furthermore, the acceleration control device 19 can control the throttle opening independently of the accelerator operation by the driver. Therefore, the vehicle control device 1 can generate any acceleration in the vehicle 100 by sending an acceleration command to the acceleration control device 19. From this, it can also be said that the vehicle control device 1 plays a role in automatically accelerating the vehicle 100 in autonomous driving where no operation by the driver is required.

[0027] [Vehicle control device configuration] Next, the configuration of the vehicle control device 1 that performs automatic driving control and is implemented in the vehicle 100 of this embodiment will be described with reference to FIG. FIG. 2 is a block diagram showing an example of the configuration of the vehicle control device 1. As shown outside the block in FIG. 2, the vehicle control device 1 is configured by, for example, a computer which is an information processing device. That is, the computer serving as the vehicle control device 1 includes a CPU (Central Processing Unit) 211 which is a processor, a memory 212, an input / output unit 213, and an interface 214.

[0028] As the memory 212, memories such as ROM (Read Only Memory) and RAM (Random Access Memory) as well as storage devices such as HDD (Hard Disk Drive) and SSD (Solid State Drive) are used. The memory 212 stores information such as programs, parameters (including thresholds), various risk maps, driving conditions, and history for operating the vehicle control device 1. Furthermore, by reading and executing programs under the control of the CPU 211, the memory 212 configures processing function units 201 to 205, which will be described later.

[0029] The input / output unit 213 performs input processing of information from the sensors 2 to 5 shown in FIG. 1 and also performs output processing of command values ​​for the actuators 10, 13, and 20. The interface 214 performs information transmission processing with other information processing devices in the vehicle 100, such as the braking control device 15, and also performs transmission and reception processing with the outside via the communication device 23 shown in FIG.

[0030] Next, we will explain the processing function units configured in the vehicle control device 1. The vehicle control device 1 is configured with an automatic driving planning unit 201, an automatic parking planning unit 202, a vehicle motion control unit 203, an actuator control unit 204, and a risk map generation unit 205. These processing function units 201 to 205 can communicate with each other via a vehicle network 206. The vehicle network 206 may be connected wirelessly as well as via a wired connection.

[0031] The automatic driving planning unit 201 plans the behavior of the vehicle 100 to automatically drive the vehicle 100 to a destination. That is, the automatic driving planning unit 201 functions as a vehicle behavior setting unit that performs a vehicle behavior setting process that sets the behavior of the vehicle, such as running or stopping. The automatic parking planning unit 202 plans the operation of the vehicle 100 to automatically park the vehicle 100 in a parking space in a parking lot or the like. The vehicle motion control unit 203 generates command values ​​for controlling the motion of the vehicle 100 during autonomous driving.

[0032] The actuator control unit 204 controls each actuator such as the engine, brake, and steering, based on the command value output from the vehicle motion control unit 203. The risk map generating unit 205 generates a risk map of vehicles and obstacles present around the host vehicle, including oncoming vehicles.

[0033] The actuator control unit 204 may be implemented in hardware different from that of the vehicle control device 1, such as an engine controller or a brake controller. The processing function units 201 to 205 may also be implemented as different arithmetic processing devices (controllers).

[0034] [Configuration of risk map generation section] Next, the configuration of the risk map generating unit 205 will be described with reference to FIG. FIG. 3 is a block diagram showing an example of the configuration of the risk map generating unit 205. As shown in FIG. The risk map generating unit 205 is supplied with information from a radar 301, a stereo camera 302, a vehicle sensor 303, and a lidar 304 as sensors that recognize the outside world.

[0035] The radar 301 emits radio waves toward an object and measures the reflected waves to measure the distance and direction to the object. Information (point cloud data) on the distance and direction to the object obtained by the radar 301 is supplied to the risk map generation unit 205. Furthermore, the stereo camera 302 can obtain depth information by simultaneously capturing images of an object from multiple different directions. The depth information obtained by the stereo camera 302 is supplied to the risk map generation unit 205.

[0036] Vehicle sensor 303 is a collective term for multiple sensors mounted on vehicle 100. In other words, vehicle sensor 303 is a group of sensors that obtain information such as vehicle speed and tire rotation speed, information calculated using GNSS (Global Navigation Satellite System) for the average position of the autonomously driven vehicle, destination information input by an occupant of the autonomously driven vehicle using a navigation system as an interface, and destination information specified by an operator or the like in a remote location using wireless communication such as a telephone line. The information obtained by these vehicle sensors 303 is supplied to risk map generation unit 205.

[0037] The lidar 304 measures scattered light in response to pulsed laser irradiation and detects the distance to a distant object. Information (point cloud data) on the distance to a distant object obtained by the lidar 304 is supplied to the risk map generation unit 205. However, the lidar 304 is not necessarily required for the risk map generation unit 205.

[0038] The risk map generation unit 205 includes a sensor information processing unit 305, a map information processing unit 306, a three-dimensional object behavior prediction unit 307, a memory unit 308, a stay risk map generation unit 309A, and a self-position estimation processing unit 310. Furthermore, the risk map generation unit 205 includes a fixed obstacle risk map generation unit 309B, a moving obstacle risk map generation unit 309C, and the like.

[0039] The sensor information processing unit 305 receives detection information from external sensors including a radar 301, a stereo camera 302, a vehicle sensor 303, and a lidar 304. The sensor information processing unit 305 obtains information about the surrounding environment from the detection information from the external sensors. The information about the surrounding environment includes, for example, object information about moving objects present around the vehicle. The sensor information processing unit 305 functions as a recognition unit that performs recognition processing to recognize the situation around the vehicle 100 based on information on the detection results of external sensors provided in the vehicle 100.

[0040] Specific object information includes attribute information of moving objects such as pedestrians, bicycles, and vehicles, as well as their current positions and current velocity vectors. Moving objects also include parked vehicles, which may move in the future even if their current velocity is zero.

[0041] The storage unit 308 stores road information and traffic light information from the point where the vehicle starts autonomous driving to the destination point and the surrounding area, route information from the current position to the destination point, a traffic rule database for the section to be traveled, etc. The storage unit 308 also stores a point cloud database used by the self-position estimation processing unit 310.

[0042] The map information processing unit 306 obtains road lane centerline information and traffic light information necessary for autonomous driving from the information stored in the memory unit 308. The map information processing unit 306 then organizes this lane centerline information and traffic light information, such as lighting information for traffic lights that the autonomous vehicle is scheduled to pass through, into information in a usable format.

[0043] Furthermore, the self-position estimation processing unit 310 estimates the location of the vehicle based on surrounding information obtained by the external sensor, the point cloud database, the steering angle of the vehicle 100, the vehicle speed, and information obtained by the GNSS. The output of the sensor information processing unit 305 , the output of the map information processing unit 306 , and the output of the self-position estimation processing unit 310 are supplied to a three-dimensional object behavior prediction unit 307 .

[0044] Based on this input information, the three-dimensional object behavior prediction unit 307 calculates future position and speed information (object prediction information) of each three-dimensional object. A three-dimensional object is a general term for moving objects, stationary objects, etc. that exist around the vehicle. To predict the movement of each moving object, the three-dimensional object behavior prediction unit 307 first predicts the position R(X(T), Y(T)) of each object at future time T based on the object information. As a specific prediction method, when the current position of the moving object is Rn0(Xn(0), Yn(0)) and the current speed is Vn(Vxn, Vyn), the three-dimensional object behavior prediction unit 307 performs prediction calculations based on, for example, the following linear prediction formula [Mathematical Expression 1].

[0045]

number

[0046] The calculation of equation (1) is performed assuming that each object moves at a constant velocity in a straight line, maintaining its current speed in the future. This allows the three-dimensional object behavior prediction unit 307 to predict many objects in a short period of time. Another calculation method is to input the position and speed information of other vehicles obtained by sensors or image information obtained by cameras into a trained neural network model. Using this method, the three-dimensional object behavior prediction unit 307 may obtain prediction results for the position and speed information of other vehicles in the future. This calculation also makes it possible to calculate the reliability of the prediction results.

[0047] The stagnation risk map generation unit 309A generates a stagnation risk map indicating the stagnation risk of surrounding vehicles, such as oncoming vehicles, based on the processing results of the sensor information processing unit 305, the map information processing unit 306, the self-position estimation processing unit 310, and the three-dimensional object behavior prediction unit 307. Collisions and deadlocks can occur when the vehicle does not behave in a coordinated manner with surrounding objects such as vehicles, pedestrians, bicycles, etc. For such scenes, the stay risk map generation unit 309A generates a stay risk map using a predicted trajectory calculated based on the target behavior of the object to behave in a coordinated manner, and outputs the generated stay risk map.

[0048] The fixed obstacle risk map generation unit 309B generates a fixed obstacle risk map that indicates the fixed obstacle risk around the vehicle based on the processing results of the sensor information processing unit 305, the map information processing unit 306, the self-position estimation processing unit 310, and the three-dimensional object behavior prediction unit 307. Note that arrows between these processing units and the fixed obstacle risk map generating unit 309B are omitted.

[0049] Similar to fixed obstacle risk map generation unit 309B, moving obstacle risk map generation unit 309C generates a moving obstacle risk map indicating the moving obstacle risk around the host vehicle based on the processing results of each processing unit.

[0050] [Stagnation risk map generation process] Next, the stagnation risk map generation process by the stagnation risk map generation unit 309A of the risk map generation unit 205 will be described with reference to FIG.

[0051] FIG. 4 is a flowchart illustrating an example of the stagnation risk map generation process. First, the risk map generating unit 205 acquires various information including peripheral information (step S401). Next, the three-dimensional object behavior prediction unit 307 of the risk map generation unit 205 predicts the target behavior of the target vehicle based on the acquired surrounding information (step S402). The prediction in step S402 is performed based on the current position and speed information of oncoming vehicles that need to pass each other in cooperation with the host vehicle, and road environment information on the road on which the host vehicle and the target vehicle are traveling. Note that the road environment information may include road shape, information on objects present in the surrounding area, map information on the surrounding area, congestion information, etc.

[0052] Then, the stay risk map generating unit 309A of the risk map generating unit 205 determines whether the target vehicle is a moving object that requires cooperative action with the host vehicle (step S403). For example, the stay risk map generating unit 309A can determine that the target vehicle is a moving object that requires cooperative action when the predicted travel trajectory of the moving object intersects with the target travel trajectory of the host vehicle, but the method of determining whether the target vehicle is a moving object that requires cooperative action is not limited to this. If it is determined in step S403 that cooperative action is necessary (YES in step S403), the stagnation risk map generation unit 309A generates and outputs a cooperative action plan and a stagnation risk map based on the target behavior of the target vehicles that require cooperative action (step S404). Moreover, if it is determined in step S403 that cooperative behavior is not necessary (NO in step S403), the stagnation risk map generating unit 309A ends the process without generating a stagnation risk map.

[0053] Next, the results of generating the collaborative action plan and congestion risk map will be explained using Figure 5. 5A and 5B show examples of the generated cooperative action plan and congestion risk map. Both Fig. 5A and Fig. 5B show an example of a situation in which a host vehicle M1 and an oncoming vehicle M2 that requires cooperative action are about to pass each other on a narrow road. The example in Fig. 5A is a case where the road boundary 741 on the left side of the host vehicle M1 is not linear but has a turnout road 2200. On the other hand, the example in Fig. 5B is a case where the road boundary 741 on the left side of the host vehicle M1 is linear and the turnout road 2200 is on the opposite side (the right side as seen from the host vehicle M1).

[0054] In this way, in the situation shown in Figures 5A and 5B, the risk map generation unit 205 predicts the target arrival point 2201 of the oncoming vehicle M2 at a future time (a predetermined time until the cooperative behavior is completed) based on various information. Then, the risk map generation unit 205 predicts the route that the oncoming vehicle M2 will take to reach the predicted destination point 2201, and calculates the area that the oncoming vehicle M2 will pass along that route. The result of the calculation by the risk map generation unit 205 becomes the retention risk map 701. The retention risk map 701 in FIG. 5A is an example in which the host vehicle M1 moves to the bypass road 2200, and the retention risk map 701 in FIG. 5B is an example in which the oncoming vehicle M2 moves to the bypass road 2200.

[0055] Furthermore, the risk map generation unit 205 calculates a host vehicle turning-off position 2202 where the host vehicle M1 should turn off, based on the destination point 2201 of the oncoming vehicle M2 and the retention risk map 701. Then, the risk map generation unit 205 considers a sequence for executing the turning-off actions in series or in parallel so that the actions of the host vehicle M1 and the oncoming vehicle M2 are both completed, and formulates a cooperative action plan for a sequence that can realize this.

[0056] As a cooperative action plan, the risk map generation unit 205 first formulates a plan for the host vehicle M1 to enter the host vehicle waiting position 2202 (step 1), then formulates a plan for the oncoming vehicle M2 to pass beside the host vehicle M1 (step 2), and finally formulates a plan for the host vehicle M1 to exit the waiting area (step 3). If a sequence that completes such cooperative behavior cannot be found, the destination point, the congestion risk map, and the vehicle's evacuation position are reviewed again, and a sequence is recursively searched for.

[0057] [Configuration of the Autonomous Driving Planning Department] Next, the configuration of the automatic driving planning unit 201 will be described with reference to FIG. FIG. 6 is a block diagram showing an example of the configuration of the automatic driving planning unit 201. The automatic driving planning unit 201 includes a driving planning unit 501 that calculates a target trajectory, a driving mode management unit 509, a trajectory planning unit 507, and a peripheral object approach determination unit 508.

[0058] The driving plan unit 501 is supplied with a delay risk map, a cooperative action plan, lane information, map information, environmental information, route information, UI (User Interface) information, etc. The driving plan unit 501 calculates weights of candidate target actions that the vehicle can take based on the route information, environmental information, detection information from the vehicle sensor 303, occupant status information, etc.

[0059] The target action candidate weights are weights for actions that the vehicle can take, such as an LK candidate weight for maintaining the lane the vehicle is currently in, an LC candidate weight for changing lanes from the current lane to an adjacent lane, an OA candidate weight for avoiding an obstacle ahead, and a CO candidate weight for cooperative action with other vehicles. For example, when traveling on a straight road, in a situation where there are no vehicles or objects ahead that need to be avoided and the route information suggests that there is no need to change lanes to an adjacent lane, the weight of the LK candidate (lane keeping) is LK = 100, the weight of the LC candidate (lane change) is LC = 0, the weight of the OA candidate (obstacle avoidance) is LC = 0, and the weight of the CO candidate (cooperative action) is CO = 0.

[0060] The trajectory planning unit 507 includes a lane keeping trajectory generation unit (LK) 502, a lane change trajectory generation unit (LC) 503, an obstacle avoidance trajectory generation unit (OA) 504, a cooperative behavior trajectory generation unit (CO) 505, and a trajectory arbitration unit 506. The trajectory planning unit 507 is an example of a trajectory planning unit that performs a traveling trajectory generation process that generates a traveling trajectory from the current position of the vehicle to a target position. The lane keeping trajectory generating unit 502 generates a trajectory for keeping the host vehicle in the center of the lane in which it is currently traveling. The lane change trajectory generating unit 503 generates a trajectory for changing lanes to an adjacent lane of the lane on which the host vehicle is currently traveling.

[0061] The obstacle avoidance trajectory generation unit 504 generates a trajectory that avoids objects that exist in the lane in which the host vehicle is currently traveling and that become obstacles to traveling. The cooperative behavior trajectory generation unit 505 generates a cooperative behavior trajectory for performing cooperative behavior with a peripheral object such as an oncoming vehicle. For example, the cooperative behavior trajectory generation unit 505 generates, as the cooperative behavior trajectory, a trajectory for moving the host vehicle to a waiting space at a waiting position where the host vehicle and a moving object (e.g., an oncoming vehicle) pass each other. The generation of the cooperative behavior trajectory takes into consideration the determination result by the peripheral object approach determination unit 508 as to whether or not an object around the host vehicle is approaching the host vehicle.

[0062] The trajectory arbitration unit 506 evaluates each of the lane-keeping trajectory, lane-changing trajectory, obstacle-avoidance trajectory, and cooperative behavior trajectory based on the degree of safety with respect to surrounding objects and the weight of the candidate target behavior, selects the trajectory with the best evaluation, and generates a target trajectory. The trajectory arbitration unit 506 can calculate the degree of safety with respect to surrounding objects based on, for example, the distance between the planned sequence of trajectory points and the surrounding objects, and the time to collision (TTC) calculated from the position and speed of the surrounding objects.

[0063] The driving mode management unit 509 calculates the previous selection information for calculating the target behavior candidate weight at the next sampling time based on the target driving mode selected by the trajectory arbitration unit 506, the trajectory evaluation value based on each target behavior candidate, and the manual driving request from the cooperative behavior trajectory generation unit 505. For example, suppose that LK (maintain the current lane: lane keep) is selected based on the evaluation values ​​of LK = 60, LC = 40, OA = 0, and CO = 0. In this case, the driving mode management unit 509 generates the previously selected current driving information (current driving mode) so that LK is more likely to be selected at the next sampling time for the sake of continuity of behavior.

[0064] [Configuration of the cooperative behavior trajectory generation unit] Next, the configuration of the cooperative behavior trajectory generation unit 505 will be described with reference to FIG. 7 is a block diagram showing an example of the configuration of the cooperative behavior trajectory generation unit 505. The cooperative behavior trajectory generation unit 505 is supplied with various types of information, such as a retention risk map, a fixed obstacle risk map, cooperative behavior plan information, environmental information, lane information, map information, and behavior prediction probability. The cooperative behavior trajectory generation unit 505 includes a retreat position and posture generation unit 601 , a cooperative behavior path generation unit 602 , and a cooperative behavior speed generation unit 603 .

[0065] The turnaround position and posture generation unit 601 generates candidate turnaround positions and postures in the collaborative action plan. The cooperative behavior path generation unit 602 calculates a travel path in the cooperative behavior mode based on the target turnaround position and posture candidates generated by the turnaround position and posture generation unit 601. The cooperative behavior speed generating unit 603 calculates the traveling speed in the cooperative behavior mode on the traveling route in the cooperative behavior mode generated by the cooperative behavior route generating unit 602.

[0066] Next, the process of generating the save position and orientation by the save position and orientation generating unit 601 will be described with reference to FIG. FIG. 8 is a flowchart illustrating an example of the retreat position and orientation generation process.

[0067] First, the turn-off position / attitude generation unit 601 acquires various information such as a retention risk map, cooperative action plan information, environmental information, lane information, map information, etc. (Step S801) Next, the turn-off position / attitude generation unit 601 generates target turn-off position / attitude candidates that represent target waiting position / attitude candidates associated with positions where the host vehicle may be present and the attitude of the host vehicle in the current cooperative action step in the cooperative action plan (Step S802).

[0068] Next, the process of generating candidate target turnaround position and orientation in step S802 will be described with reference to FIG. FIG. 9 is a diagram illustrating an example of a process for generating candidate target turnaround positions and attitudes. In the example of Figure 9, on a road with a narrow width and an area where both the host vehicle M1 and the oncoming vehicle M2 cannot or have difficulty passing through at the same time, there is an oncoming vehicle M2 ahead of the host vehicle M1 that should act cooperatively. There is also a waiting space ahead and to the left of the host vehicle M1. In other words, the waiting space ahead and to the left of the host vehicle M1 is a passable area where the host vehicle M1 and the oncoming vehicle M2 can pass through at the same time. In the example of Figure 9, the passable area refers to an area within the drivable area defined by the road boundary 741 where the host vehicle M1 and the oncoming vehicle M2 can pass through at the same time if either vehicle withdraws to the waiting space.

[0069] The stay risk map 701 generated by the stay risk map generating unit 309A (FIG. 3) is a risk map that exists ahead of the host vehicle M1 and on the road surface on which the host vehicle M1 is located. Therefore, in order to realize passing driving, which is a cooperative behavior, the host vehicle M1 needs to move to a location where the stay risk map 701 does not exist and wait there. Therefore, the stay risk map generating unit 309A generates turn-off position and attitude candidates (N1, θ1) to (N5, θ5) in the turn-off space, which is a surrounding drivable area (inside the road boundary 741) where the stay risk map 701 does not exist. N represents the planar position coordinate, and θ represents the headway angle of the vehicle. For example, the headway angle is the angle between the direction of the road and the vehicle. The number of turn-off position and attitude candidates generated at this time is arbitrary. As the location for generating the turn-off position and attitude candidates, a location close to the stay risk map 701 can be selected within a range that takes into account the size of the host vehicle M1.

[0070] [Cooperative behavior trajectory generation processing] Next, the collaborative behavior path generation process performed by the collaborative behavior path generation unit 602 will be described with reference to the flowchart of FIG. FIG. 10 is a flowchart illustrating an example of a collaborative behavior path generation process.

[0071] First, the cooperative action path generation unit 602 acquires various information such as a stay risk map, cooperative action plan information, environmental information, lane information, map information, and peripheral object approach determination information (step S1001). Next, the cooperative behavior path generating unit 602 generates arrival position candidates, which are arrival positions at which the traveling mode is changed from the cooperative behavior mode to another mode after passing the turning-off position (step S1002). Thereafter, the cooperative behavior path generating unit 602 generates cooperative behavior trajectory candidates that represent candidates for cooperative behavior trajectories that pass through the turn-off position candidates and the arrival position candidates (step S1003).

[0072] Furthermore, the cooperative behavior path generating unit 602 evaluates the generated cooperative behavior trajectory candidates and selects the optimal cooperative behavior trajectory (step S1004). In this embodiment, the selected cooperative behavior trajectory is called a "cooperative behavior path" and is used as the basis for the cooperative behavior trajectory when approaching a peripheral object that is generated by the peripheral object approach trajectory generating unit 605. Finally, the cooperative behavior path generating unit 602 stores information on the cooperative behavior trajectory (cooperative behavior path) selected in step S1004 in the storage unit 308 (memory 212 in FIG. 2) (step S1005).

[0073] The process of evaluating the generated cooperative behavior route candidates and selecting the optimal cooperative behavior route in step S1004 will be described below. To select the most appropriate cooperative behavior route from the generated cooperative behavior route candidates, it is possible to use, for example, the equation in [Mathematical Expression 2] as a method of calculating and evaluating each cooperative behavior route candidate. As described above, cooperative behavior route candidates are generated for each of the candidate evacuation positions and candidate arrival positions, and the cooperative behavior route generation unit 602 must select the most appropriate one from among these candidates. Therefore, for each cooperative behavior route candidate, the following evaluation function is used to determine an overall evaluation value for each cooperative behavior route candidate:

[0074]

number

[0075] In formula [2], a weighting coefficient (i) is set for each of the five individual evaluation values, and the sum of the products of the weighting coefficient (i) and the individual evaluation value (i) is calculated. The weighting coefficient is a value determined in advance depending on the scene. The smaller the overall evaluation value, the more preferable the trajectory candidate is. The five individual evaluation values ​​are an evaluation of safety, convenience, ride comfort, discomfort, and pressure on oncoming vehicles.

[0076] Safety is evaluated by assessing whether a sufficient distance is maintained to avoid contact with or approaching nearby objects while traveling along a passing route. To this end, the cooperative behavior path generation unit 602 calculates a safety evaluation value using the distance to nearby objects and a fixed obstacle risk value (not shown). For example, the evaluation value can be the reciprocal of the smallest distance between the road boundary and the route candidate, or the value (probability) of a collision risk map (not shown). In this embodiment, the route is represented by planar coordinate information, and the trajectory includes time information in addition to route information.

[0077] Convenience is evaluated by evaluating cooperative behavior in situations where an oncoming vehicle is present. For example, if it takes a long time for the vehicle to pass an oncoming vehicle, the oncoming vehicle will have to wait longer, which may lead to distrust among other vehicles and the occupants of the vehicle. Therefore, the evaluation value of convenience is calculated using the length of the vehicle's evacuation route, assuming that the vehicle's speed is constant and does not depend on the length of the route. For example, one possible method is to use the length of the route on the entrance side (the route leading into the evacuation area) as the evaluation value.

[0078] The evaluation of ride comfort is based on whether the acceleration and time derivative of acceleration that occur in the vehicle when the vehicle follows the evacuation route become large. For example, a route with a large lateral acceleration and a long duration for which lateral acceleration occurs is considered to have a poor ride comfort. Also, when the turning radius of the route is small, the larger the total turning angle, the less comfortable the route becomes. Therefore, it is possible to use the value of (reciprocal of turning radius) x (total turning angle) as an evaluation index.

[0079] The evaluation of discomfort is based on the number of times that the occupants of the vehicle have to turn around while passing other vehicles. In other words, it is considered that turning around while passing other vehicles causes discomfort, and the evaluation value of discomfort is calculated using the number of times that the occupants of the vehicle have to turn around while passing other vehicles.

[0080] In evaluating the sense of pressure felt by oncoming vehicles, a route in which the subject vehicle approaches too closely to an oncoming vehicle may cause a sense of pressure that may cause distrust or fear among the occupants of the subject vehicle and the other vehicle. Therefore, an evaluation value for the sense of pressure felt by oncoming vehicles is calculated based on the distance between the subject vehicle's candidate cooperative behavior route and the oncoming vehicle. For example, a method can be considered in which the reciprocal of the smallest distance between the candidate cooperative behavior route and the center position coordinates of the other vehicle is used as an evaluation index for the sense of pressure.

[0081] [Generation of arrival position candidates] Next, a method for generating arrival position candidates in step S1002 will be described with reference to Fig. 11. Fig. 11 is a diagram showing examples of arrival position candidates.

[0082] Here, a point where a transition from the cooperative behavior mode to another mode (for example, lane keeping mode) is expected is selected as the arrival position candidate. In the example of FIG. 11, in a situation where a location where passing is possible (for example, a turning-off space) is set by a road boundary 741, a position on the center line 1101 of the lane of travel of the host vehicle M1 stored in the map information, which is away from the turning-off position / posture candidate point P2 in the traveling direction of the host vehicle by the distance between the cooperative behavior mode start point P1 and the turning-off position / posture candidate point P2 (a position where the length of L in FIG. 11 is equal), can be set as the arrival position candidate point P3. Note that the method of generating the arrival position candidate is not limited to this.

[0083] [Generation of collaborative action route candidates] Next, a method for generating collaborative action route candidates in step S1003 will be described with reference to FIG. FIG. 12 is a diagram showing an example of a cooperative behavior route (passing route). One possible method for generating a cooperative behavior path is to use the aforementioned congestion risk map and candidate evacuation positions and postures to generate arcs, straight lines, and clothoid curves for the target candidate evacuation positions and postures and candidate arrival positions.

[0084] Fig. 12 shows the result of generating a target route for the host vehicle M1 from its current position to the waiting position using straight lines and arc curves. In the example of Fig. 12, the route for the host vehicle M1 from its current position to the waiting position is a straight line that is approximately parallel to the center line 1101 at the linear road boundary 741 until the waiting position begins. After that, at the point where the road boundary 741 deviates to the outside of the road, the route toward the waiting position (N3, θ3) becomes a curved route using arc curves. Then, after the host vehicle M1 passes the oncoming vehicle M2, the route returns to the center line 1101.

[0085] Next, we will explain the cooperative behavior speed generation unit 603. In general, when calculating a target speed profile for traveling on a target route, a method can be considered in which routes that satisfy the variational equation of the following [Equation 3] are selected as candidates.

[0086]

number

[0087] Here, one possible method for calculating the future motion state of the host vehicle on its path is to use a plant model of the vehicle, such as a bicycle model or a four-wheel model. This plant model makes it possible to obtain the behavior of the vehicle (longitudinal acceleration, lateral acceleration, headway angle, etc.) when traveling along a target route. As the output of this equation, it is possible to generate multiple s velocity profiles by changing the weights (w1, w2, w3) of each coefficient or by changing the items to be evaluated.

[0088] Next, the congestion risk map and cooperative behavior trajectory when there is an oncoming vehicle ahead of the host vehicle will be explained using the schematic diagram of a typical scene shown in Figure 13. FIG. 13 shows an example of a typical scene of a congestion risk map and a cooperative behavior trajectory when an oncoming vehicle M2 is present ahead of the host vehicle M1.

[0089] Here, as shown in Fig. 13A, the host vehicle M1 and the oncoming vehicle M2, which is the target of cooperative behavior, are traveling with a passing point between them. There is a waiting area on the left side of the host vehicle M1. In addition, a delay risk map 701 calculated based on the surrounding environment and the location of the oncoming vehicle M2 is acquired. In this situation, as shown in Fig. 13B, a waiting position 702 for the host vehicle M1, a trajectory (route) 704 from the current position to the waiting position 702, and a trajectory (route) 705 from the waiting position 702 to a candidate arrival position 703 are calculated based on the delay risk map 701.

[0090] [When the vehicle's occupants feel uneasy or uncomfortable] Next, referring to FIG. 14, a case will be described in which an oncoming vehicle or the surrounding environment causes a feeling of anxiety or discomfort to the occupants of the vehicle when traveling along the above-mentioned cooperative behavior trajectory (following the planned cooperative behavior trajectory).

[0091] Fig. 14 is a diagram showing an example in which an occupant of the host vehicle feels uneasy or uncomfortable when traveling along a cooperative behavior trajectory. As shown in Fig. 14, there is a waiting space 1400 on the host vehicle's side and a waiting space 1401 in the lane on the oncoming vehicle's side, both of which have spaces available for waiting. Based on the behavior prediction probability (execution probability) by the three-dimensional object behavior prediction unit 307 of the host vehicle, it is predicted that the oncoming vehicle M2 will most likely take shelter in the waiting space 1401. Therefore, a retention risk map 701 is generated that indicates that the oncoming vehicle M2 will enter the waiting space 1401. As described above, possible prediction methods include linear prediction and methods using statistical models such as neural networks, but other methods may also be used.

[0092] 14A, the host vehicle M1 is attempting to achieve cooperative behavior by following trajectories 704 and 705 on the cooperative behavior trajectory. At this time, during the course of the predicted behavior of the oncoming vehicle, the oncoming vehicle was initially predicted to be taking a turn at time T=T0, but as a result of a new prediction at time T=T2, the oncoming vehicle does not take a turn at turn-off space 1401 but continues toward the host vehicle. At this stage, the most likely prediction result is that an oncoming vehicle is moving toward the host vehicle. Therefore, the congestion risk map at time T=T1 is generated as shown in Figure 14B, and as a result, the host vehicle M1 needs to take cooperative action to retreat to the waiting space 1400 on its own vehicle side.

[0093] In this way, even if a congestion risk map is generated and a cooperative action plan is formulated based on the most likely prediction results at the initial time, it is possible that the driving intentions of oncoming vehicles may change, or detection or prediction errors of oncoming vehicles may make it impossible to carry out the cooperative action originally envisioned.

[0094] In such a case, the host vehicle M1 needs to generate a backward trajectory 704 and retreat while backing up, as shown in Fig. 14B. Backing up generally makes the occupants feel uncomfortable or scared, and is expected to result in a lower traveling speed than when traveling forward. On the other hand, in such a case, the occupants of the vehicle M1, the occupants of the oncoming vehicle M2, and surrounding traffic participants may feel uncomfortable and traffic flow may be disrupted, which may have a negative impact on surrounding traffic, such as causing congestion.

[0095] It is usually desirable to create a driving plan that does not interfere with the driving of the oncoming vehicle M2 and is robust against changes in the behavioral intention of the oncoming vehicle and changes in the environment. The vehicle control device 1 of this embodiment plans a turning-off position and a trajectory to the turning-off position in order to reduce the fear felt by occupants when the host vehicle M1 and surrounding vehicles act in coordination, even when passing each other is possible. As a result, the vehicle control device 1 can cause the host vehicle M1 to pull over to a location where passing can be achieved in a safer and more secure state, and can prevent the occupants from feeling uneasy.

[0096] For this reason, in this embodiment, as shown in Fig. 15, the cooperative behavior trajectory generation unit 505 includes a cooperative behavior speed branch point search calculation unit 1501 and a cooperative behavior speed profile calculation unit 1502. The cooperative behavior speed profile calculation unit 1502 performs processing to search for speed branch points in the cooperative behavior speed branch point search calculation unit 1501, and generates a speed profile during cooperative behavior based on the speed branch points. Note that the speed branch points can be considered as specific positions, and the cooperative behavior trajectory generation unit 505 functions as a specific position setting unit that performs processing to set the specific positions.

[0097] [Cooperative Action Speed ​​Branch Point Search Processing] Next, the process of searching for a speed branch point (specific position) by the cooperative behavior speed branch point search calculation unit 1501 (specific position setting unit) will be described with reference to FIG. FIG. 16 is a diagram showing an example of speed branch points calculated by the cooperative behavior speed branch point search calculation unit 1501 (specific position setting unit).

[0098] First, the three-dimensional object behavior prediction unit 307, which is a behavior estimation unit, performs a behavior estimation process to generate a behavior prediction of a moving object other than the host vehicle (for example, an oncoming vehicle). The three-dimensional object behavior prediction unit 307 estimates at least one or more behavior candidates (behavior patterns) of the oncoming vehicle. For example, in the scene of FIG. 16, the behavior predicted as behavior candidate number 1 is that the oncoming vehicle will take shelter in the waiting space 1401 on the oncoming lane side. Furthermore, the behavior candidate number 2 is that the oncoming vehicle will proceed straight ahead as a result of the driver of the oncoming vehicle or the cruise control device (for autonomous driving) assuming that the host vehicle will take shelter in the waiting space 1400.

[0099] Here, it is assumed that the respective behavior prediction probabilities are 60% for behavior candidate number 1 and 40% for behavior candidate number 2. In this case, the predicted probability of behavior candidate number 1 is the highest behavior probability, and it becomes behavior probability rank 1. And behavior candidate number 2 becomes behavior probability rank 2. Note that the behavior prediction probability is a reliability, and can also be considered as a likelihood. In this example, two behavior candidates are used, but the solid object behavior prediction unit 307 may predict three or more behavior candidates (behavior patterns).

[0100] Next, the waiting position and attitude generation unit 601 determines a waiting position 702 based on the behavior candidate (behavior pattern) associated with behavior candidate number 1, which has the highest behavior probability. The waiting position 702 is a target position for the host vehicle M1 to move to a passable area on a road where there are areas that are impassable or difficult for the host vehicle M1 and the oncoming vehicle M2 to pass at the same time. The passable area is an area where the host vehicle M1 and the oncoming vehicle M2 can pass at the same time. "Impassable" refers to, for example, a state where the road width is narrower than the combined width of the host vehicle M1 and the oncoming vehicle M2, making it impossible for the two vehicles to pass each other, or a state where the host vehicle M1 and the oncoming vehicle M2 cannot pass each other due to the presence of an obstacle (such as a utility pole or a parked vehicle). "Impassable" refers to, for example, a state where the road width is wider than the combined width of the host vehicle M1 and the oncoming vehicle M2, but the vehicles get too close to each other when passing each other, causing anxiety for the occupants or hindering driving. In the example of FIG. 16, the passable area refers to an area within the drivable area that has a road width that allows the host vehicle M1 and the oncoming vehicle M2 to pass through simultaneously due to the escape spaces 1400 and 1401.

[0101] Next, the cooperative behavior path generation unit 602, which is a path generation unit, determines a path 704 for the host vehicle M1 to a retreat position 702 within the passable area for at least one candidate behavior (behavior pattern) of the oncoming vehicle M2 estimated by the three-dimensional object behavior prediction unit 307.

[0102] Next, a speed branch point (specific position) P11 is set on this route 704. The cooperative behavior speed branch point search calculation unit 1501 (specific position setting unit) can set at least one speed branch point (specific position) P11 on the route 704 from the current position of the host vehicle M1 to the evacuation position 702 based on the behavior prediction probability of the oncoming vehicle M2. As an example of a method for setting the speed branch point P11, a method for setting a speed branch point based on the behavior prediction probability and the distance L to the evacuation position 702 will be described. The cooperative behavior speed branch point search calculation unit 1501 can calculate the product of the distance L to the evacuation position 702 and a first behavior prediction probability (for example, a maximum behavior probability that is the maximum behavior prediction probability among multiple behavior prediction probabilities), and set a point at that distance from the evacuation position 702 as the speed branch point P11. 16, the cooperative behavior speed branch point search calculation unit 1501 can set a point obtained by dividing the road distance L from the current vehicle position in a ratio of 40:60 as the speed branch point P11. In other words, the cooperative behavior speed branch point search calculation unit 1501 can calculate the product of the first behavior prediction probability with respect to the distance of the route 704 from the current position of the vehicle M1 to the waiting position, and set a position that is the product value away from the waiting position as the specific position. Note that the method for setting the speed branch point (specific position) P11 is not limited to this example. Other methods for setting the speed branch point (specific position) P11 will be described later.

[0103] Next, the cooperative behavior speed profile calculation unit 1502 will be described with reference to FIG. 17. The cooperative behavior speed profile calculation unit 1502 functions as a host vehicle behavior setting unit that sets the speed of the host vehicle M1. The cooperative behavior speed profile calculation unit 1502 (host vehicle behavior setting unit) sets the speed to the speed branch point (specific position) set by the cooperative behavior speed branch point search calculation unit 1501 (specific position setting unit) as described below. For example, it is desirable that the target speed from the current position of the host vehicle M1 to the speed branch point (specific position) be equal to or lower than the normal driving section speed set for the road on which the host vehicle M1 is traveling. The normal driving section speed can be, for example, the legal speed set for the road on which the host vehicle M1 is traveling.

[0104] By doing this, even if the behavioral intention of the oncoming vehicle M2 changes (for example, if the behavioral intention of the oncoming vehicle M2 changes from behavior candidate number 1 to behavior candidate number 2), the host vehicle M1 can extend the time it takes to reach the specific position because the host vehicle M1 is traveling at a speed slower than the normal traveling section speed. In other words, the host vehicle M1 can stay longer in the traveling section where it can retreat to the host vehicle retreat space 1400 without having to travel backward. In other words, by extending the time it takes the host vehicle M1 to reach the specific position, it can maintain a state where it can retreat to the host vehicle retreat space 1400 by traveling forward until the behavioral intention of the oncoming vehicle M2 is determined. As a result, the host vehicle M1 can retreat to the host vehicle retreat space 1400 more frequently without having to travel backward.

[0105] Alternatively, the cooperative behavior speed profile calculation unit 1502 can set the target speed from the current vehicle position to a specific position to a speed equal to or lower than the product of a first behavior prediction probability (e.g., the most likely behavior prediction probability) and the normal driving section speed set for the road on which the vehicle M1 is traveling, based on the behavior prediction probability of the oncoming vehicle M2.

[0106] By doing this, even if the behavioral intention of the oncoming vehicle changes (for example, if the behavioral intention of the oncoming vehicle changes from candidate behavior number 1 to candidate behavior number 2), the vehicle can travel at a slower speed than when the behavioral intention of the oncoming vehicle remains unchanged from candidate behavior number 1. Here, a high probability of a change from candidate behavior number 1 to candidate behavior number 2 means that the difference between the highest probability and the second highest probability is small. This increases the frequency with which the host vehicle M1 can retreat into the host vehicle retreat space 1400 without having to travel backward.

[0107] Alternatively, the target speed of the host vehicle M1 from the current position to the specific position may be set at the speed according to the method described above, and when the host vehicle M1 reaches the vicinity of the specific position, if the most likely behavior prediction probability of the oncoming vehicle M2 is smaller than a predetermined value, the host vehicle M1 may decelerate to a speed at which the host vehicle M1 can stop at the specific position and stop. In other words, the cooperative behavior speed profile calculation unit 1502 can set the speed of the host vehicle M1 so that the host vehicle M1 decelerates toward the specific position and stops at the specific position. This process of decelerating to a speed at which the host vehicle M1 can stop at the specific position and stop can also be applied when the target speed from the current position to the specific position is equal to or less than the product of a first behavior prediction probability (e.g., the most likely behavior prediction probability) and a normal driving section speed set for the road on which the host vehicle is traveling, based on the behavior prediction probability of the moving object.

[0108] Even when the host vehicle M1 reaches the specific position, the behavioral intention of the oncoming vehicle M2 may change, and if the host vehicle M1 passes the specific position at the normal driving section speed, it may be necessary to reverse in order to retreat to the retreat position 702. Therefore, if the behavior prediction probability of the oncoming vehicle M2's most likely behavior is below a predetermined value set in advance, it is desirable to stop the host vehicle M1 at the specific position and confirm the retreat behavior of the host vehicle M1 after the behavior prediction probability of the oncoming vehicle M2 becomes sufficiently high (for example, when it exceeds a predetermined value or a predetermined time has passed). This increases the frequency at which the host vehicle M1 can retreat to the host vehicle retreat space 1400 without retreating. Finally, the generated cooperative behavior trajectory (for passing) is stored in the memory unit 308.

[0109] [Configuration of the driving mode management unit] Next, the driving mode management unit 509 (FIG. 6) will be described with reference to FIG. FIG. 18 is a block diagram showing an example of the configuration of the driving mode management unit 509. Various information such as a target driving mode, a manual driving request, and system information is supplied to the driving mode management unit 509. The system information is, for example, information on the power switch of the vehicle 100 and information on the automatic driving switch of the automatic driving system.

[0110] 18, the driving mode management unit 509 manages three states: a manual driving mode 1801, a system-off mode 1802 of the automatic driving system, and an automatic driving mode 1803. The driving mode management unit 509 outputs the current mode state as the current driving mode.

[0111] When the power switch or the like of the vehicle 100 is turned on (an example of UI information in FIG. 6), the driving mode management unit 509 transitions from the system off mode 1802 to the manual driving mode 1801 or the like. In addition, the driving mode management unit 509 transitions from the manual driving mode 1801 or the like to the automatic driving mode 1803 when the driver or the like turns on the automatic driving switch (an example of UI information in FIG. 6).

[0112] As already explained, the autonomous driving mode 1803 includes lane keeping (LK) and lane change (LC). Intermediate transition states exist between these states. In the cooperative behavior mode (CO), a deadlock or an event requiring a change of driving direction to the driver may occur, and the autonomous driving system may determine that it is necessary to take over (give up). In such cases, the cooperative behavior trajectory selection unit 606 (FIG. 7) issues a manual driving request, resulting in a transition to the manual driving mode 1801.

[0113] According to this embodiment, when cooperative driving with a nearby moving object is performed, a speed branch point (specific position) is set based on the predicted behavior probability of the oncoming vehicle, and the speed of the own vehicle up to the speed branch point (specific position) is set, thereby enabling robust driving against changes in the behavioral intention of the oncoming vehicle. Therefore, even if a change occurs in the behavioral intention of the oncoming vehicle, cooperative driving is possible without the need to perform actions such as backing up. This reduces the possibility of the occupants of the vehicle M1, the oncoming vehicle M2, and other traffic participants feeling uncomfortable, or of adversely affecting surrounding traffic, such as causing congestion due to disruption of traffic flow.

[0114] <Second embodiment> Next, a vehicle control device according to a second embodiment of the present invention will be described with reference to FIGS. In the second embodiment, the configuration of the vehicle control device and the configuration of the vehicle equipped with the vehicle control device are the same as those described in the first embodiment, and therefore, redundant explanations will be omitted. When describing each component of the vehicle control device, the same reference numerals as those described in Figures 1 to 18 of the first embodiment will be used.

[0115] The vehicle control device 1 of this embodiment is desired to create a driving plan that does not interfere with the driving of the oncoming vehicle M2 and is robust against changes in the behavioral intention of the oncoming vehicle and changes in the environment. Here, the vehicle control device 1 of this embodiment plans a turning-off position and a trajectory to the turning-off position in order to reduce the fear felt by occupants when the host vehicle M1 and surrounding vehicles act in coordination, even though passing each other is possible. Unlike the first embodiment, the vehicle control device 1 of this embodiment searches for a route branching point based on the shape of a candidate evacuation route for a candidate evacuation location calculated based on multiple behavior prediction results, and adjusts the speed to that route branching point, thereby reducing the discomfort described above.

[0116] For this purpose, in this embodiment, in the cooperative behavior trajectory generation unit 505 in FIG. 19, a cooperative behavior path branch point search calculation unit 1901 and a cooperative behavior speed profile calculation unit 1902 generate a speed profile during cooperative behavior.

[0117] [Route branch point search processing in the cooperative action route branch point search calculation unit] Next, the process of searching for route branch points by the collaborative behavior route branch point search calculation unit 1901 will be described with reference to Fig. 20. The collaborative behavior route branch point search calculation unit 1901 functions as a specific position setting unit that searches for route branch points and sets the route branch points as specific positions. 20 is a diagram showing an example of route branch points calculated by the cooperative behavior route branch point search calculation unit 1901. In this scene, as in the first embodiment, two possible predicted behavior candidates of the oncoming vehicle M2 estimated by the behavior estimation unit are a route 1601 in which the oncoming vehicle M2 takes shelter and a route 1602 in which the oncoming vehicle M2 proceeds straight. Here, it is assumed that the behavior pattern in which the oncoming vehicle M2 takes shelter on the route 1601 is the most likely behavior prediction probability, and the behavior pattern in which the oncoming vehicle M2 proceeds straight on the route 1602 is the second most likely behavior prediction probability. Note that, although two behavior patterns are estimated as the behavior patterns of the oncoming vehicle M2 here, three or more behavior patterns may be estimated.

[0118] First, the cooperative behavior path generation unit 602, which is a path generation unit, generates turn-off positions 702A and 702B for each of the behavior prediction results (paths 1601 and 1602), as shown in Fig. 20A. Then, the cooperative behavior path generation unit 602, which is a path generation unit, generates paths 704A and 704B to the turn-off positions 702A and 702B. Here, the cooperative behavior path branch point search calculation unit 1901 (specific position determination unit) extracts a common portion that can be shared between the two paths, calculates the position where the common portion ends, starting from the current position of the vehicle M1, as a path branch point, and can set the path branch point as a specific position. In other words, the cooperative behavior path branch point search calculation unit 1901 (specific position determination unit) can set a specific position at a position where a path 704A determined for a behavior pattern associated with a first behavior prediction probability (e.g., the most likely behavior prediction probability) of the oncoming vehicle M2 intersects with a path 704B determined for a behavior pattern associated with a second behavior prediction probability (e.g., the second most likely behavior prediction probability) of the oncoming vehicle M2. Note that although paths are generated for two behavior patterns of the oncoming vehicle M2 here, paths may be generated for three or more behavior patterns of the oncoming vehicle M2. In this case, a specific position can be set at a position where the paths generated for the three or more behavior patterns intersect.

[0119] However, the route to the waiting position may also be a route that moves forward once and then reverses to enter waiting position 702B, as shown in Figure 20B. In a case like Figure 20B, the route branch point is created farther away from the current vehicle position. However, such a route 704B includes a reverse route and is therefore not considered to be an optimal waiting route.

[0120] Therefore, it is desirable that the cooperative behavior route branch point search calculation unit 1901 (specific position determination unit) evaluates each route using the method described above and selects the optimal route. However, a method of changing the optimal route within a predetermined range to share it with other candidate actions is also possible. In this way, the speed profile calculation unit 1902 calculates the speed profile from the current position to the specific position and from the specific position to the evacuation position based on the route branch points (specific positions) calculated by the cooperative behavior route branch point search calculation unit 1901.

[0121] In other words, when there are multiple waiting positions 702A and 702B, the cooperative action route branch point search calculation unit 1901 (specific position determination unit) can find routes 704A and 704B that allow the host vehicle M1 to reach each of the multiple waiting positions 702A and 702B without moving in a direction opposite to the direction of travel of the host vehicle M1 (backing up), and set the route branch points where the respective routes 704A and 704B branch off as speed branch points (specific positions).

[0122] 21 shows an example of a calculation of a speed profile from the current position to the route branching point and the turning-off position based on the route branching point. As in the first embodiment, the traveling speed up to the route branching point is preferably equal to or less than the normal traveling section speed. Alternatively, the traveling speed may be equal to or less than the product of the most likely behavior prediction probability of the oncoming vehicle M2 and the normal traveling section speed set for the road on which the host vehicle M1 is traveling.

[0123] By traveling at such a speed, even if the behavioral intention of the oncoming vehicle M2 changes before the route branch point, it becomes possible to take shelter at another shelter position, for example, a shelter position based on the behavior with the second highest behavior prediction probability, without having to make troublesome route changes such as backing up. This reduces the possibility that the occupants of the vehicle M1, the oncoming vehicle M2, and other traffic participants will feel uncomfortable, and that adverse effects on surrounding traffic, such as the occurrence of congestion due to traffic flow disruption, can be reduced.

[0124] <Modification> The present invention is not limited to the above-described embodiments, and various other applications and modifications are possible without departing from the spirit of the present invention as set forth in the claims. For example, the above-described embodiments have been described in detail and specifically to clearly explain the present invention, and are not necessarily limited to those including all of the components described. Furthermore, it is also possible to add, replace, or delete other components to or from part of the configuration of each embodiment.

[0125] For example, the passing behavior described in each figure is an example of passing by pulling your vehicle over to the left side of the road in areas where vehicles drive on the left side of the road, and in areas where vehicles drive on the right side of the road, the left and right directions when passing are reversed.

[0126] Furthermore, the passing operation described in FIG. 5 and the like is an example in which there is a turnout road 2200 that protrudes from the boundary of the road, but this is just one example. For example, as shown in FIG. 22, the present invention is also applicable to a case where a vehicle M1 cannot pass an oncoming vehicle M2 due to obstacles 2110, 2120 such as utility poles on the road.

[0127] In the example of Fig. 22, as shown in Fig. 22A, one example is when standby position candidate 1, which is an internal standby position, is set between two obstacles 2101 and 2102. Another example is when standby position candidate 2, which is an external standby position, is set further in front of the front obstacle 2101, as shown in Fig. 22B. Here, as explained in the above-mentioned embodiment example, the vehicle control device 1 can deal with this by selecting the optimal waiting position candidate and speed for the vehicle M1 based on the behavior prediction probability of the oncoming vehicle M2, which is a moving body.

[0128] The present invention is also applicable when the host vehicle or an oncoming vehicle passes each other during automatic parking assistance. For example, the example shown in FIG. 23A is a processing example when the host vehicle M1, approaching the oncoming vehicle M2, detects that the oncoming vehicle M2 is performing automatic parking assistance to park in a parking space 2301. In this case, the vehicle control device 1 of the host vehicle M1 generates a retention risk map 2302 that assumes the host vehicle M1 will move forward diagonally in front of the parking space 2301 and then move backward from that position to the parking space 2301, in response to the oncoming vehicle M2's driving for automatic parking. Then, the vehicle control device 1 generates a waiting position candidate 2210 (waiting position candidate 1) at a position that does not overlap with the retention risk map 2302, and sets a speed for moving toward the waiting position candidate 1.

[0129] The example shown in FIG. 23B is a case where the host vehicle M1 is performing automatic parking assistance for the parking space 2301. In this case, the vehicle control device 1 of the host vehicle M1 generates a retention risk map 2303 for traveling at a location away from the parking space 2301 of the oncoming vehicle M2. Then, the vehicle control device 1 generates a candidate waiting position 2220 (candidate waiting position 2) at the side of the parking space 2301 at a position that does not overlap with the retention risk map 2303, and sets a speed for moving toward the candidate waiting position 2.

[0130] In this way, when the vehicle M1 comes face to face with the oncoming vehicle M2 near the parking space 2301, the vehicle control device 1 of the vehicle M1 calculates the predicted behavior probability of the oncoming vehicle M2, selects which of the two candidate waiting positions 2210, 2220 to head to, and sets the speed.

[0131] Furthermore, as a process for when there are obstacles 2110, 2120 such as utility poles on the road as described in FIG. 22, the vehicle M1 may move to a position where one of the waiting position candidates is retreated, as shown in FIG. That is, the vehicle control device 1 of the host vehicle M1 sets a candidate waiting position 2101 (interior waiting: upper part of Fig. 24) between obstacles 2110 and 2120 ahead of the host vehicle M1, and a candidate waiting position 2103 (backing up waiting: lower part of Fig. 24) behind the host vehicle M1, as shown in Fig. 24. Then, the vehicle control device 1 of the host vehicle M1 selects which of the two candidate waiting positions 2210, 2220 to head towards based on the predicted behavior probability of the oncoming vehicle M2, and sets the speed.

[0132] In addition, in each of the above-described embodiments, a driving trajectory (cooperative behavior trajectory) is generated (calculated), and the vehicle is driven autonomously along the driving trajectory. Generally, a driving trajectory includes information on a coordinate position on a road and the time at which the vehicle passes through the coordinate position, and in the present invention, driving is controlled using information on the coordinate position and time. In contrast, the driving trajectory may be information on a so-called driving route that includes only information on the coordinate position on a road, without information on the time at which the vehicle passes each coordinate position.

[0133] When passing each other, it is preferable to predict the position of the oncoming vehicle at each time and determine where the vehicle will move at each time. However, in situations where the oncoming vehicle is stopped, it is also possible to generate a driving route that contains only coordinate position information instead of a driving trajectory. By generating a travel route in this manner, it is not necessary to handle time information, and the travel trajectory (travel route) can be generated more easily.

[0134] Furthermore, in the configuration shown in FIG. 2, the vehicle control device 1 configured as a computer is configured as a device that performs the processing of each of the above-mentioned embodiments, but a program implemented in an existing vehicle control device 1 may be modified to perform similar processing. In this case, the program may be stored and transferred in a recording medium such as an external memory, an IC card, an SD card, an optical disk, or the like, in addition to being prepared in the memory within the computer shown in Fig. 2. Furthermore, part or all of the vehicle control device 1 may be realized by dedicated hardware such as an FPGA (Field Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit).

[0135] In the above-described embodiment, the control lines and information lines of the vehicle control device and each processing function unit are shown as necessary for explanation, and not all control lines and information lines are necessarily shown in the product. In reality, it may be considered that almost all components are interconnected. Also, with regard to the flowcharts shown in Figures 4, 8, and 10, the processing order may be changed or multiple processes may be executed simultaneously as long as the processing results are the same.

[0136] Furthermore, in each of the above-described embodiments, examples have been described in which vehicles pass each other, but the oncoming vehicle may be in a form other than a vehicle as long as it is a mobile body capable of autonomous movement. Furthermore, although the term "parallel" is used in this specification, this term does not mean only "parallel" in the strict sense, but also includes the meaning of "parallel" in the strict sense and also includes the meaning of "approximately parallel" within the range in which the function can be exerted. [Explanation of symbols]

[0137] 1...vehicle control device, 100...vehicle, 201...automatic driving planning unit, 202...automatic parking planning unit, 203...vehicle motion control unit, 204...actuator control unit, 205...risk map generation unit, 211...CPU, 212...memory, 213...input / output unit, 214...interface, 301...radar, 302...stereo camera, 303...vehicle sensor, 304...lidar, 305...sensor information processing unit, 306...map information processing unit, 307...solid object behavior prediction unit, 308...memory unit, 309A...stay risk map generation unit, 309B...fixed obstacle risk map generation unit, 309C...moving obstacle risk map generation unit, 310...self-position estimation processing unit, 501...driving planning unit, 502...lane keeping trajectory generation unit, 503...lane change trajectory generation unit, 504... Obstacle avoidance trajectory generation unit, 505... Cooperative behavior trajectory generation unit, 506... Trajectory arbitration unit, 507... Trajectory planning unit, 508... Peripheral object approach determination unit, 509... Driving mode management unit, 601... Evacuation position and attitude generation unit, 602... Cooperative behavior path generation unit, 603... Cooperative behavior speed generation unit, 1501... Cooperative behavior speed branch point search calculation unit (specific position setting unit), 1502... Cooperative behavior speed profile calculation unit, 1801... Manual driving mode, 1802... Autonomous driving system off mode, 1803... Autonomous driving mode, 1901... Cooperative behavior path branch point search calculation unit, 1902... Speed ​​profile calculation unit when route branch point is used

Claims

1. a behavior estimation unit that estimates at least one behavior pattern of a moving object on a road on which the host vehicle and a moving object other than the host vehicle are present; a path generation unit that defines an area through which the host vehicle and the moving object can pass simultaneously as a passable area, and generates a path for the host vehicle to a target position within the passable area based on at least one behavior pattern of the moving object estimated by the behavior estimation unit; a specific position setting unit that sets at least one specific position on the route from the current position of the vehicle to the target position; and a host vehicle behavior setting unit that sets a speed of the host vehicle in a section from a current position of the host vehicle to the specific position based on a behavior prediction probability that indicates the possibility that the moving object will perform the behavior pattern estimated by the behavior estimation unit. Vehicle control device.

2. The specific position setting unit sets at least one specific position on a route from the current position of the vehicle to the target position based on the behavior prediction probability. The vehicle control device according to claim 1 .

3. The specific position setting unit calculates a product of the first behavior prediction probability and a distance of the route from the current position of the vehicle to the target position on the route determined for a behavior pattern related to a first behavior prediction probability of the moving object, and sets the specific position at a position that is a distance from the target position by the product value. The vehicle control device according to claim 1 .

4. The host vehicle behavior setting unit sets a target speed from the current position of the host vehicle to the specific position to a speed lower than a normal driving section speed set for a road on which the host vehicle is traveling. The vehicle control device according to claim 1 .

5. The host vehicle behavior setting unit sets a speed of the host vehicle so that the host vehicle decelerates toward the specific position and stops at the specific position when the most likely behavior prediction probability of the moving object is smaller than a predetermined value. The vehicle control device according to claim 1 .

6. The host vehicle behavior setting unit sets a target speed from the current position of the host vehicle to the specific position to a speed equal to or less than a product of a most likely behavior prediction probability of the moving object and a normal driving section speed set for the road on which the host vehicle is traveling. The vehicle control device according to claim 1 .

7. When a plurality of target positions exist, the specific position setting unit determines a route by which the host vehicle can reach each of the plurality of target positions without moving in a direction opposite to the traveling direction of the host vehicle, and sets a branch point of the route as the specific position. The vehicle control device according to claim 1 .

8. The specific position setting unit sets the specific position at a position where the route determined for one behavior pattern of the mobile object intersects with the route determined for another behavior pattern of the mobile object. The vehicle control device according to claim 1 .

9. A vehicle control method in which a control of a host vehicle is performed by a processing device, a behavior estimation process for estimating at least one behavior pattern of the moving object on a road on which the vehicle and the moving object are present; a route generation process in which an area through which the host vehicle and the moving object can simultaneously pass is defined as a passable area, and the host vehicle generates a route to a target position within the passable area for at least one behavior pattern of the moving object estimated by the behavior estimation process; a specific position setting process for setting at least one specific position on the route from the current position of the vehicle to the target position; and a host vehicle behavior setting process for setting a speed of the host vehicle in a section from a current position of the host vehicle to the specific position based on a behavior prediction probability indicating the possibility that the moving object will perform the behavior pattern estimated by the behavior estimation process. Vehicle control method.

Images