Automatic driving path generating device and automatic driving device

Abstract

An autonomous driving travel path generation device identifies a mobile object located around a vehicle, calculates the vehicle's autonomous driving own vehicle path and multiple predicted paths for the mobile object based on the vehicle's position on a map, the mobile object's position on a map, and map information, calculates, for each predicted path, a predicted acceleration generated by the mobile object moving along the predicted path, determines, based on a comparison result of the predicted acceleration with an acceleration threshold, a predicted path used in generating the travel path, i.e., an object path, from among the multiple predicted paths, and generates a travel path based on the own vehicle path and the object paths.

Description

Technical Field

[0001] The present invention relates to an automatic driving path generation device and an automatic driving device. Background Art

[0002] Japanese Patent Application Laid-Open No. 2008-158969 is a well-known technical document related to the generation of autonomous driving routes. This publication describes a device that calculates the probability of a collision between a vehicle and another vehicle by finding paths that interfere with the vehicle's path among predicted paths that other vehicles can take. Summary of the Invention

[0003] When multiple predicted paths for a moving object are calculated based on its position on a map and map information, the moving object may not actually move along the predicted path, depending on its speed and other behaviors, resulting in a low probability of achieving the predicted path. When the predicted path for the moving object is used to generate the path of the vehicle itself (the driver's own vehicle), it is believed that such a low probability predicted path will have a minimal impact. Therefore, it is desirable to generate a path for autonomous driving of the vehicle based on the probability of achieving the predicted path.

[0004] In the present technical field, it is desirable to provide an autonomous driving route generation device and an autonomous driving device that can appropriately generate an autonomous driving route based on the feasibility of a predicted route of a moving object located around a vehicle.

[0005] One technical solution of the present invention is an autonomous driving travel path generation device for generating an autonomous driving travel path for a vehicle, comprising: a moving object recognition unit that recognizes a moving object located around the vehicle; a path calculation unit that calculates the vehicle's own vehicle path for autonomous driving and multiple predicted paths for the moving object based on the vehicle's position on a map, the moving object's position on a map, and map information; a predicted acceleration calculation unit that calculates, for each predicted path, a predicted acceleration generated by the moving object moving along the predicted path based on the multiple predicted paths and the speed of the moving object; an object path determination unit that determines a predicted path used in generating the travel path, i.e., an object path, from among the multiple predicted paths based on a comparison result of the predicted acceleration with an acceleration threshold; and a travel path generation unit that generates a travel path based on the own vehicle path and the object paths.

[0006] According to one technical solution of the present invention, an autonomous driving path generation device calculates a predicted acceleration of a moving object for each predicted path based on multiple predicted paths and the speed of the moving object. Based on the comparison result of the predicted acceleration with an acceleration threshold, a predicted path used in path generation, namely, an object path, is determined. A path is generated based on the own vehicle path and the object path. In this way, by using the object path determined based on the comparison result of the predicted acceleration with the acceleration threshold, a path can be generated that takes into account the feasibility of the moving object's predicted path. This makes it possible to more appropriately generate an autonomous driving path compared to a case where the feasibility of the moving object's predicted path is not taken into account.

[0007] In an autonomous driving path generation device according to one aspect of the present invention, the predicted acceleration calculation unit may calculate, for each predicted path, a predicted lateral acceleration generated by a mobile body moving along the predicted path based on the curvature radius of the predicted path and the speed of the mobile body, and the target path determination unit may determine, as the target path, a predicted path whose maximum value of the predicted lateral acceleration among the plurality of predicted paths is less than a lateral acceleration threshold. In this case, the autonomous driving path can be generated more appropriately by utilizing the fact that the probability of achieving a predicted path, such as one in which the maximum value of the predicted lateral acceleration generated by the mobile body is greater than the lateral acceleration threshold, is low.

[0008] In an autonomous driving route generation device according to one aspect of the present invention, the predicted acceleration calculation unit may calculate, for each predicted route, a predicted longitudinal acceleration generated by a mobile body moving along the predicted route based on the mobile body's temporary stopping position or deceleration position on the predicted route and the mobile body's speed, and the target route determination unit may determine, as the target route, a predicted route for which the maximum magnitude of the predicted longitudinal acceleration among the plurality of predicted routes is less than a longitudinal acceleration threshold. In this case, the autonomous driving route can be generated more appropriately by utilizing the fact that the predicted route for which the maximum magnitude of the predicted longitudinal acceleration generated by the mobile body is greater than the longitudinal acceleration threshold is considered to have a low probability of being achieved.

[0009] In an autonomous driving route generation device according to one aspect of the present invention, the map information may include priority information for determining traffic regulatory priorities for a plurality of vehicles, and the target route determination unit may further determine, as the target route, a predicted route for a mobile object whose priority among the plurality of predicted routes is higher than the priority of the vehicle. In this case, the autonomous driving route can be generated more appropriately by utilizing the fact that the predicted route, which is considered to have a lower probability of being realized, is lower than the priority of the vehicle.

[0010] In an autonomous driving route generation device according to one aspect of the present invention, the moving object recognition unit may recognize a direction indication of the moving object, and the target route determination unit may further determine, as the target route, a predicted route extending in the direction indicated by the direction indication from among the plurality of predicted routes. In this case, the autonomous driving route can be generated more appropriately by utilizing the fact that the predicted route extending in the direction not indicated by the direction indication is considered to have a low probability of being achieved.

[0011] In an autonomous driving path generation device according to one aspect of the present invention, the moving object recognition unit may recognize a lateral position shift of the moving object in a direction intersecting the moving direction, and the target path determination unit may further determine, as the target path, a predicted path extending in a direction corresponding to the lateral position shift from among the plurality of predicted paths. In this case, the autonomous driving path can be generated more appropriately by utilizing the fact that the predicted path extending in a direction not corresponding to the lateral position shift is considered to have a low probability of being realized.

[0012] Another technical solution of the present invention is an automatic driving device for making a vehicle automatically drive according to a generated travel path, comprising: a moving body recognition unit for recognizing moving bodies located around the vehicle; a path calculation unit for calculating the vehicle's own vehicle path for automatic driving and a plurality of predicted paths for the moving body based on the vehicle's position on a map, the moving body's position on a map, and map information; a predicted acceleration calculation unit for calculating, for each predicted path, a predicted acceleration generated by the moving body moving along the predicted path based on a plurality of predicted paths and the speed of the moving body; and an object path determination unit for determining the object path based on the predicted acceleration and the acceleration. A predicted path used in generating a travel path, i.e., a target path, is determined based on a comparison result of speed thresholds; a travel path generation unit generates a travel path based on the own vehicle path and the target path; and a driving control unit causes the vehicle to automatically drive according to the travel path. When the own vehicle path and the target path interfere with each other, the travel path generation unit calculates an interference position between the own vehicle path and the target path, and generates an interference-avoiding travel path as a travel path that causes the vehicle to stop before the interference position based on the vehicle speed, the own vehicle path, and the interference position. The driving control unit decelerates the vehicle according to the interference-avoiding travel path.

[0013] According to another aspect of the present invention, an automated driving device calculates a predicted acceleration of a moving object for each predicted path based on multiple predicted paths and the speed of the moving object. Based on the comparison result of the predicted acceleration with an acceleration threshold, a target path, which is a predicted path used in generating a travel path, is determined from among the predicted paths. A travel path is generated based on the own vehicle path and the target path. By using the target path determined based on the comparison result of the predicted acceleration with the acceleration threshold, a travel path can be generated that takes into account the feasibility of the moving object's predicted path. This allows for more appropriate automated driving travel path generation compared to a case where the feasibility of the moving object's predicted path is not considered. Furthermore, according to another aspect of the present invention, an automated driving device generates an interference-avoidance travel path when the own vehicle path interferes with the target path, and decelerates the vehicle according to the interference-avoidance travel path. Consequently, the vehicle stops before the interference location. In this way, by decelerating the vehicle while taking into account the feasibility of the moving object's predicted path, the vehicle can be decelerated more appropriately compared to a case where the feasibility of the moving object's predicted path is not considered.

[0014] According to some technical solutions of the present invention, it is possible to appropriately generate an autonomous driving travel path based on the feasibility of a predicted path of a moving object located around a vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Features, advantages, and technical and industrial significance of exemplary embodiments of the present invention will be described below with reference to the accompanying drawings, in which like reference numerals represent like elements, and wherein:

[0016] Figure 1 This is a block diagram illustrating an automatic driving device including the automatic driving route generation device according to the first embodiment.

[0017] Figure 2A This is a schematic top view illustrating a predicted path of an oncoming vehicle at an intersection.

[0018] Figure 2B This is an example of Figure 2A A schematic top view of the predicted path of an oncoming vehicle traveling at a high speed at an intersection.

[0019] Figure 3 It shows Figure 1 A flowchart of an example of autonomous driving processing by an ECU.

[0020] Figure 4 It shows Figure 3 This is a flowchart of an example of the processing of S05 and S06.

[0021] Figure 5 It shows Figure 3This is a flowchart of an example of the processing of S08 and S09.

[0022] Figure 6 This is a block diagram illustrating an automatic driving device including an automatic driving route generation device according to a second embodiment.

[0023] Figure 7A This is a schematic plan view illustrating a predicted path of a merging vehicle at a Y-junction.

[0024] Figure 7B This is an example of Figure 7A A schematic top view of the predicted path of a merging vehicle traveling at a low speed at a Y-intersection.

[0025] Figure 8 It shows Figure 6 A flowchart of an example of autonomous driving processing by an ECU.

[0026] Figure 9 It shows Figure 8 This is a flowchart of an example of the processing of S35 and S36.

[0027] Figure 10 This is a flowchart showing an example of additional target path determination processing.

[0028] Figure 11 This is a flowchart showing another example of the additional object path determination process.

[0029] Figure 12 This is a flowchart showing another example of the additional object path determination process. DETAILED DESCRIPTION

[0030] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

[0031] [First embodiment]

[0032] Figure 1The autonomous driving path generation device 20 of the first embodiment shown is a device that is mounted on a vehicle and generates an autonomous driving path for the vehicle. The autonomous driving path generation device 20 constitutes a part of the autonomous driving device 100 that causes the vehicle to automatically drive according to the generated path. The autonomous driving path of the vehicle includes the path the vehicle is traveling and the speed of the vehicle. Here, as the autonomous driving path, it is assumed that the vehicle will travel a path within a few seconds to a few minutes when a moving object other than the vehicle is located around the vehicle on the road on which the vehicle (the vehicle itself) is traveling. The moving object refers to a dynamic obstacle located around the vehicle. Examples of the moving object include other vehicles, bicycles, pedestrians, etc. As an example, the moving object here is another vehicle. Other vehicles can include various vehicles such as passenger cars, trucks, buses, motorcycles, bicycles, etc.

[0033] For example, autonomous driving refers to a driving state in which a vehicle automatically drives along a road. This includes a driving state in which the vehicle automatically drives toward a predetermined destination without the driver performing any steering operation. Autonomous driving includes levels 2 to 4 of SAE (Society of Automotive Engineers) J3016.

[0034] Hereinafter, the configuration of the automatic driving route generating device 20 and the automatic driving device 100 according to the first embodiment will be described with reference to the accompanying drawings. Figure 1 As shown, the automated driving route generation device 20 includes an ECU (Electronic Control Unit) 10, which serves as a unified management system. The ECU 10 is an electronic control unit that includes a CPU (Central Processing Unit), ROM (Read Only Memory), RAM (Random Access Memory), and a CAN (Controller Area Network) communication circuit. In the ECU 10, various functions are achieved by, for example, loading programs stored in the ROM into the RAM and executing them on the CPU. The ECU 10 may also be composed of multiple electronic units.

[0035] The ECU 10 is connected to a GPS receiving unit 1 , an external sensor 2 , an internal sensor 3 , and a map database 4 .

[0036] The GPS receiving unit 1 receives signals from three or more GPS satellites to measure the position of the vehicle (for example, the latitude and longitude of the vehicle). The GPS receiving unit 1 transmits the measured vehicle position information to the ECU 10 .

[0037] External sensor 2 is a device that detects conditions surrounding the vehicle. External sensor 2 includes at least one of a camera and a radar sensor. External sensor 2 can also be configured to reconstruct various properties of the vehicle's external environment (such as the vehicle's position, relative distance from a moving object, relative speed from a moving object, moving object's orientation, lane shape, and signal light status).

[0038] The camera is a device that captures the vehicle's exterior. It is located behind the vehicle's windshield. The camera transmits image information related to the vehicle's exterior to the ECU 10. The camera can be either a monocular or a stereo camera. A stereo camera has two imaging units configured to reproduce binocular parallax. The image information from the stereo camera also includes depth information.

[0039] A radar sensor is a detection device that uses radio waves (such as millimeter waves) or light to detect moving objects around the vehicle. Radar sensors include, for example, millimeter-wave radars or laser radars (LiDAR). Radar sensors detect moving objects and stationary objects by transmitting radio waves or light to the vehicle's periphery and receiving the radio waves or light reflected by the moving objects. The radar sensor transmits information about the detected moving objects to the ECU 10. Stationary objects include utility poles, buildings, traffic lights, and the like. Stationary objects may also include white lines.

[0040] Internal sensors 3 are devices that detect the vehicle's driving state. They include a vehicle speed sensor, an acceleration sensor, and a yaw rate sensor. The vehicle speed sensor detects the vehicle's speed. For example, a wheel speed sensor, installed on a vehicle wheel or a drive shaft that rotates integrally with the wheel, can be used to detect the wheel's rotational speed. The vehicle speed sensor transmits the detected vehicle speed information (wheel speed information) to the ECU 10.

[0041] An acceleration sensor detects the acceleration of the vehicle. Examples of acceleration sensors include a longitudinal acceleration sensor that detects the vehicle's longitudinal acceleration and a lateral acceleration sensor that detects the vehicle's lateral acceleration. The acceleration sensor transmits vehicle acceleration information to the ECU 10. A yaw rate sensor detects the yaw rate (rotational angular velocity) of the vehicle's center of gravity about a vertical axis. For example, a gyroscope sensor can be used as the yaw rate sensor. The yaw rate sensor transmits the detected vehicle yaw rate information to the ECU 10.

[0042] The map database 4 is a database that stores map information. The map database 4 is stored, for example, in a storage medium such as an HDD (Hard Disk Drive) mounted on the vehicle. Map information includes information on road locations, road shape (e.g., types of curves and straight sections, radius of curvature of curves, shape of intersections, lane widths, etc.), location information on intersections and forks, and location information on structures. Road shape information can be, for example, a straight or curved line connecting points (e.g., nodes) arranged at the center of a lane. Structures include facilities such as shops located along the road. Map information includes information such as speed limits (e.g., the legal maximum speed) corresponding to the location of lanes on the map. The speed limit can be any speed below the legal maximum speed. Map information may also include various traffic regulation information associated with the location of lanes on the map (e.g., information on sections where lane changes are permitted and sections where lane changes are not permitted). Furthermore, the map database 4 may be stored in a computer at a facility such as a management center that can communicate with the vehicle.

[0043] The actuator 5 is a device used in the automatic driving control of the vehicle. The actuator 5 includes at least a driving actuator, a braking actuator and a steering actuator. The driving actuator controls the air supply to the engine (throttle opening) according to the control signal from the ECU 10, thereby controlling the driving force of the vehicle. In addition, when the vehicle is a hybrid vehicle, in addition to the air supply to the engine, a control signal from the ECU 10 is input to the motor serving as the power source to control the driving force. When the vehicle is an electric vehicle, a control signal from the ECU 10 is input to the motor serving as the power source to control the driving force. The motor serving as the power source in these cases constitutes the actuator 5.

[0044] The brake actuator controls the braking system based on control signals from the ECU 10, controlling the braking force applied to the vehicle's wheels. For example, a hydraulic brake system can be used as the braking system. The steering actuator controls the driving of the assist motor in the electric power steering system, which controls the steering torque, based on control signals from the ECU 10. Thus, the steering actuator controls the steering torque of the ECU 10.

[0045] Next, the functional configuration of the ECU 10 will be described. The ECU 10 includes a vehicle position recognition unit 11, an external environment recognition unit (mobile object recognition unit) 12, a driving state recognition unit 13, a path calculation unit 14, a predicted acceleration calculation unit 15, a target path determination unit 16, a travel path generation unit 17, and a driving control unit 18. The autonomous driving path generation device 20 includes at least the external environment recognition unit 12, the path calculation unit 14, the predicted acceleration calculation unit 15, the target path determination unit 16, and the travel path generation unit 17.

[0046] The vehicle position recognition unit 11 identifies the vehicle's position on a map based on the position information from the GPS receiver 1 and the map information in the map database 4. Furthermore, the vehicle position recognition unit 11 utilizes the position information of fixed objects such as utility poles included in the map information in the map database 4 and the detection results of the external sensor 2 to identify the vehicle's position using SLAM (Simultaneous Localization and Mapping) technology. Alternatively, the vehicle position recognition unit 11 may identify the vehicle's position on a map using known methods.

[0047] The vehicle exterior environment recognition unit 12 recognizes the vehicle exterior environment based on the detection results of the external sensor 2. The vehicle exterior environment recognition unit 12 recognizes the vehicle exterior environment based on the camera's image information and the radar sensor's moving object information using a known method.

[0048] The vehicle exterior environment recognition unit 12 identifies mobile objects located around the vehicle based on the detection results of the external sensor 2. The vehicle exterior environment includes the position of the mobile object relative to the vehicle, the relative speed of the mobile object relative to the vehicle, and the direction of movement of the mobile object relative to the vehicle. The vehicle exterior environment recognition unit 12 identifies the speed of the mobile object based on the vehicle speed and the relative speed of the mobile object. The vehicle exterior environment recognition unit 12 identifies the location of the mobile object on the map based on the position of the mobile object relative to the vehicle and map information. The vehicle exterior environment recognition unit 12 may also assign a mobile object identification number, etc., to each mobile object in response to a mobile object signal representing the mobile object.

[0049] The driving state recognition unit 13 identifies the state of the driving vehicle based on the detection results of the internal sensor 3. The driving state includes the vehicle's speed, acceleration, and yaw rate. Specifically, the driving state recognition unit 13 identifies the vehicle's speed based on speed information from the speed sensor. The driving state recognition unit 13 identifies the vehicle's acceleration (fore-aft acceleration and lateral acceleration) based on acceleration information from the acceleration sensor. The driving state recognition unit 13 identifies the vehicle's yaw rate based on yaw rate information from the yaw rate sensor.

[0050] The path calculation unit 14 calculates the vehicle's own vehicle path for autonomous driving and multiple predicted paths for the mobile body based on the vehicle's position on the map, the mobile body's position on the map, and the map information. The vehicle's own vehicle path for autonomous driving is the path of the vehicle used to automatically drive the vehicle along the vehicle's target route. The predicted path of the mobile body is the path that the mobile body is likely to travel from the current position of the mobile body. The predicted path can be, for example, a path within the range that the mobile body will travel within a few seconds to a few minutes. The path calculation unit 14 can also calculate the vehicle's own vehicle path and the predicted path after a preset set time (for example, after 1 second, after 3 seconds, etc.).

[0051] The path calculation unit 14 calculates the vehicle's potential risk area based on, for example, information from the GPS receiver 1, external sensors 2, internal sensors 3, and the map database 4, and the target route. The path calculation unit 14 can calculate the vehicle's path by performing a valley search on the potential risk area using a known method.

[0052] The path calculation unit 14 calculates the potential risk area for the vehicle using a known method based on, for example, the vehicle's speed, location, width of the road on which the vehicle is traveling, and the location of one or more targets. Targets include fixed and mobile obstacles detected in front of the vehicle. Fixed obstacles include, for example, road surface paint (including lane boundary lines such as white and yellow lines) and structures (such as curbs, pillars, utility poles, buildings, signs, and trees). Mobile obstacles include, for example, pedestrians, bicycles, strollers, and other vehicles.

[0053] The path calculation unit 14 calculates a predicted path for the mobile object based on, for example, map information and the vehicle's external environment. The predicted path for the mobile object may include multiple paths corresponding to the road environment surrounding the mobile object. For example, the path calculation unit 14 uses the shape of a preceding intersection on the road on which the mobile object is traveling, parking lots of stores or other facilities located near the mobile object, as destinations, and calculates one or more paths that the mobile object is likely to take from its current location as predicted paths.

[0054] Figure 2A This is a schematic top view illustrating a predicted path of an oncoming vehicle at an intersection. Figure 2B This is an example of Figure 2A A schematic top view of the predicted path of an oncoming vehicle traveling at a high speed at an intersection. Figure 2A In FIG, another vehicle V2 is shown as an example of a moving body, which is an oncoming vehicle relative to the own vehicle V1. Figure 2B In, as a comparison Figure 2A The oncoming vehicle is another vehicle V3 traveling at a high speed.

[0055] like Figure 2A and Figure 2B As shown, the own vehicle V1, other vehicles V2 and other vehicles V3 are about to arrive at the intersection RX. Figure 2A and Figure 2B In the example shown in FIG. 1 , the path calculation unit 14 calculates a vehicle path P1a as the vehicle path for autonomous driving of the vehicle V1 based on the position of the vehicle V1 on the map and the map information. The vehicle path P1a is a straight path for the vehicle V1 from the road R1 to the road R3.

[0056] exist Figure 2A In the example shown in FIG. 1 , the path calculation unit 14 calculates predicted paths P2a, P2b, and P2c as predicted paths for the other vehicle V2. Predicted path P2a is the straight path for the other vehicle V2 from road R3 to road R1. Predicted path P2b is the left turn path for the other vehicle V2 from road R3 to road R4. Predicted path P2c is the right turn path for the other vehicle V2 from road R3 to road R2.

[0057] exist Figure 2B In the example shown in FIG. 1 , the path calculation unit 14 calculates predicted paths P3a, P3b, and P3c as predicted paths for the other vehicle V3. Predicted path P3a is the straight path for the other vehicle V3 from road R3 to road R1. Predicted path P3b is the left turn path for the other vehicle V3 from road R3 to road R4. Predicted path P3c is the right turn path for the other vehicle V3 from road R3 to road R2.

[0058] Here, the predicted path of the moving body includes paths that are physically unlikely to be realized based on the speed of the moving body and other behaviors, and paths that cannot be traveled in accordance with traffic regulations that the moving body should comply with. As described later, since these paths are not determined as target paths, they are not used in generating the vehicle's travel path. Figure 2A and Figure 2B In the example, other vehicle V3 is traveling at a higher speed than other vehicle V2. Other vehicle V2's speed allows it to travel along predicted paths P2b and P2c. However, other vehicle V3's speed is so high that it would be difficult for it to actually travel along predicted paths P3b and P3c. Therefore, predicted paths P3b and P3c are unlikely to be realized.

[0059] The predicted acceleration calculation unit 15 calculates, for each predicted path, the predicted acceleration of the moving object as it moves along the predicted path based on the multiple predicted paths and the speed of the moving object. The predicted acceleration is the acceleration predicted to be generated by the moving object as it moves along the predicted path. The predicted acceleration varies depending on the speed of the moving object.

[0060] The predicted acceleration calculation unit 15 may also identify the type of moving object (e.g., passenger car, truck, bus, motorcycle, bicycle, pedestrian, etc.) and calculate the predicted acceleration based on the type of moving object using a pre-prepared motion model. The predicted acceleration calculation unit 15 may also calculate the predicted acceleration of the moving object using various well-known techniques related to vehicle and pedestrian behavior prediction. The predicted acceleration calculation unit 15 may also limit the moving objects to be predicted to those on the road. In this case, the predicted acceleration calculation unit 15 may also perform the behavior prediction calculation assuming the moving object is a vehicle.

[0061] The predicted acceleration can include predicted lateral acceleration and predicted longitudinal acceleration. The predicted lateral acceleration is the lateral acceleration predicted to be generated by the moving body when the moving body moves along the curved predicted path. The lateral direction refers to the vehicle width direction of the moving body.

[0062] The predicted acceleration calculation unit 15 here calculates the predicted lateral acceleration as the predicted acceleration. The predicted acceleration calculation unit 15 calculates the predicted lateral acceleration generated by the moving body moving along the predicted path for each predicted path based on the curvature radius of the predicted path and the speed of the moving body. Figure 2A In the example of , the predicted acceleration calculation unit 15 calculates the centrifugal force generated by the other vehicle V2 as the predicted lateral acceleration based on the curvature radius of the predicted path in the curved section of the predicted paths P2b and P2c and the speed of the other vehicle V2. Figure 2B In the example of , the predicted acceleration calculation unit 15 calculates the centrifugal force generated by the other vehicle V3 as the predicted lateral acceleration based on the curvature radius of the predicted path in the curved sections of the predicted paths P3b and P3c and the vehicle speed of the other vehicle V3.

[0063] The target path determination unit 16 determines a target path based on the comparison result between the predicted acceleration and the acceleration threshold. The target path is a predicted path among the multiple predicted paths used to generate the travel path. The target path determination unit 16 determines as the target path the predicted path whose maximum predicted lateral acceleration among the multiple predicted paths is less than the lateral acceleration threshold.

[0064] The lateral acceleration threshold is a threshold of predicted lateral acceleration used to determine whether the predicted path of a moving object is a path with a low probability of being physically realized. The lateral acceleration threshold can also be a pre-set parameter or mapping. The lateral acceleration threshold can also be set based on the radius of curvature of the predicted path. For example, the lateral acceleration threshold can be set so that the smaller the radius of curvature of the predicted path, the larger the threshold. Furthermore, the lateral acceleration threshold can also be set based on the type of moving object.

[0065] Specifically, the target path determination unit 16 may also determine whether the maximum value of the predicted lateral acceleration is less than a lateral acceleration threshold. If the target path determination unit 16 determines that the maximum value of the predicted lateral acceleration is less than the lateral acceleration threshold, the predicted path is determined as the target path. If the target path determination unit 16 determines that the maximum value of the predicted lateral acceleration is greater than the lateral acceleration threshold, the predicted path is not determined as the target path. Furthermore, "not determined as the target path" includes, for example, when evaluating multiple paths using probability values ​​corresponding to potential risk areas, where the probability values ​​are sufficiently small that the predicted path is not substantially determined.

[0066] exist Figure 2A In the example of , the target path determination unit 16 determines that the maximum value of the predicted lateral acceleration for the predicted paths P2a, P2b, and P2c is less than the lateral acceleration threshold. The target path determination unit 16 determines the predicted paths P2a, P2b, and P2c as the target paths.

[0067] exist Figure 2B In the example shown in FIG. 1 , the target path determination unit 16 determines that the maximum value of the predicted lateral acceleration for predicted path P3a is less than the lateral acceleration threshold, and determines that the maximum values ​​of the predicted lateral acceleration for predicted paths P3b and P3c are greater than the lateral acceleration threshold. The target path determination unit 16 determines predicted path P3a as the target path. On the other hand, the target path determination unit 16 does not determine predicted paths P3b and P3c as the target path.

[0068] The route generation unit 17 generates a route based on the vehicle's own route and the target route. The route generation unit 17 generates a route such that the vehicle and the moving object do not collide with each other, for example.

[0069] As an example, the path generation unit 17 determines whether the own vehicle path and the object path interfere with each other. When the path generation unit 17 determines that the own vehicle path interferes with the object path, it generates a first path that avoids interference (interference-avoiding path). The first path is a path that does not cause the vehicle and the moving object to collide with each other. The first path can be set as a path that stops the vehicle before the interference position, for example. When the own vehicle path interferes with the object path, the path generation unit 17 calculates the interference position between the own vehicle path and the object path. The path generation unit 17 generates the first path based on the vehicle speed, the own vehicle path and the interference position. The path generation unit 17 generates the first path, for example, according to a speed plan that decelerates the vehicle traveling along the own vehicle path and stops before the interference position.

[0070] If the path generation unit 17 determines that the vehicle's own path and the target path do not interfere, it generates a second path. The second path is a path for the vehicle that does not involve an avoidance maneuver with respect to the moving object. The second path can be, for example, a path for the vehicle to travel along the vehicle's own path.

[0071] The driving control unit 18 automatically drives the vehicle according to the path generated by the path generation unit 17. The driving control unit 18 controls the actuator 5 so that the vehicle travels according to the path generated by the path generation unit 17. For example, if the driving control unit 18 determines that the vehicle's path interferes with the target path, the driving control unit 18 controls the vehicle's travel according to a first path, causing the vehicle to stop before the interference point. If the driving control unit 18 determines that the vehicle's path does not interfere with the target path, the driving control unit 18 controls the vehicle's travel according to a second path, causing the vehicle to travel along the vehicle's path.

[0072] exist Figure 2A In the example, the predicted path P2c among the predicted paths P2a, P2b, and P2c determined as the target path interferes with the own vehicle path P1a. As a result, the ECU 10 determines that the own vehicle path interferes with the target path. The travel path generation unit 17 obtains the coordinates on the map of the position where the own vehicle path P1a and the predicted path P2c intersect as the interference position. The travel path generation unit 17 obtains the interference time based on the interference position between the own vehicle path P1a and the predicted path P2c, the position on the map of the own vehicle V1, and the speed of the own vehicle V1. The travel path generation unit 17 generates a first travel path according to a speed plan such that the own vehicle V1 traveling along the own vehicle path P1a decelerates and stops before the interference position. The travel control unit 18 controls the actuator 5 so that the own vehicle V1 stops before the interference position according to the first travel path.

[0073] exist Figure 2B In the example shown in FIG. 1 , predicted path P3a, which has been determined as the target path, does not interfere with the host vehicle path P1a, and predicted paths P3b and P3c have not been determined as target paths. Therefore, the ECU 10 determines that the host vehicle path does not interfere with the target path. The path generation unit 17 generates a second path along the host vehicle path P1a. The travel control unit 18 controls the actuator 5 so that the host vehicle V1 travels along the second path along the host vehicle path P1a.

[0074] Next, the processing of the automatic driving route generating device 20 and the ECU 10 of the automatic driving device 100 will be described with reference to the drawings.

[0075] Reference Figure 3 The processing of the ECU 10 will be described. Figure 3 It shows Figure 1 Flowchart of an example of automatic driving processing by ECU 10. Figure 3 The flowchart shown is executed, for example, during automated driving.

[0076] like Figure 3 As shown, as S01, the automatic driving route generation device 20 and the ECU 10 of the automatic driving device 100 recognize the vehicle's position on a map via the vehicle position recognition unit 11. The vehicle position recognition unit 11 recognizes the vehicle's position on a map based on the position information of the GPS receiving unit 1 and the map information in the map database 4.

[0077] In S02, the ECU 10 recognizes the driving state of the vehicle through the driving state recognition unit 13. The driving state recognition unit 13 recognizes at least the vehicle speed as the driving state of the vehicle based on the detection result of the internal sensor 3.

[0078] In S03, the ECU 10 uses the vehicle exterior environment recognition unit 12 to identify the moving object. Based on the detection results of the external sensor 2, the vehicle exterior environment recognition unit 12 recognizes the vehicle exterior environment and the moving object. For example, the vehicle exterior environment recognition unit 12 identifies the position of the moving object relative to the vehicle, the relative speed of the moving object relative to the vehicle, and the direction of movement of the moving object relative to the vehicle as the vehicle exterior environment. The vehicle exterior environment recognition unit 12 identifies the speed of the moving object based on the vehicle speed and the relative speed of the moving object. Based on the position of the moving object relative to the vehicle and map information, the vehicle exterior environment recognition unit 12 identifies the location of the moving object on the map.

[0079] In S04, the ECU 10 calculates a plurality of predicted routes of the moving object through the route calculation unit 14. The route calculation unit 14 calculates a plurality of predicted routes of the moving object based on the position of the moving object on the map and the map information. Figure 2A In the example of , the path calculation unit 14 calculates predicted paths P2a, P2b, and P2c as a plurality of predicted paths.

[0080] In S05, the ECU 10 calculates the predicted acceleration for each predicted path through the predicted acceleration calculation unit 15. The predicted acceleration calculation unit 15 calculates the predicted acceleration generated by the moving body moving along the predicted path for each predicted path based on the multiple predicted paths and the speed of the moving body. In S06, the ECU 10 determines the target path through the target path determination unit 16. The target path determination unit 16 determines the target path based on the comparison result of the predicted acceleration and the acceleration threshold. In S05 and S06, the ECU 10 specifically performs Figure 4 The processing shown.

[0081] Figure 4 It shows Figure 3 This is a flowchart of an example of the processing of S05 and S06. Figure 4 The flowchart shown is used, for example, in autonomous driving Figure 1 ECU processing is executed. Figure 4 When there are multiple prediction paths, the processing may be repeatedly performed on each prediction path.

[0082] like Figure 4 As shown, the ECU 10 performs the process of S11 for each predicted path as the process of S05 described above. In S11, the ECU 10 calculates the predicted lateral acceleration for each predicted path based on the curvature radius of the predicted path and the speed of the mobile object via the predicted acceleration calculation unit 15. For example, the predicted acceleration calculation unit 15 calculates the centrifugal force generated by the mobile object as the predicted lateral acceleration based on the curvature radius of the predicted path and the speed of the mobile object during the curved sections of the predicted paths P2b and P2c.

[0083] The ECU 10 performs steps S12 through S14 for each predicted path as the aforementioned step S06. In S12, the ECU 10 uses the target path determination unit 16 to determine whether the maximum value of the predicted lateral acceleration is less than the lateral acceleration threshold. If the ECU 10 determines that the maximum value of the predicted lateral acceleration is less than the lateral acceleration threshold (S12: Yes), the process proceeds to S13. If the ECU 10 determines that the maximum value of the predicted lateral acceleration is greater than the lateral acceleration threshold (S12: No), the process proceeds to S14.

[0084] In S13, the ECU 10 determines the predicted route as the target route via the target route determination unit 16. On the other hand, in S14, the ECU 10 does not determine the predicted route as the target route via the target route determination unit 16.

[0085] After the above-mentioned processing of S13 or S14, the ECU 10 ends Figure 4 processing, move to Figure 3 Processing of S07.

[0086] In S07, the ECU 10 calculates the vehicle's own vehicle path through the path calculation unit 14. The path calculation unit 14 calculates the vehicle's own vehicle path for automatic driving based on the vehicle's position on the map and map information. Figure 2A and Figure 2B In the example of , the path calculation unit 14 calculates the own vehicle path P1a as the own vehicle path.

[0087] In S08, the ECU 10 generates a path for the automatic driving of the vehicle through the path generation unit 17. The path generation unit 17 generates a path based on the own vehicle path and the target path. In S09, the ECU 10 controls the driving of the vehicle through the driving control unit 18. The driving control unit 18 controls the actuator 5 so that the vehicle drives according to the path generated by the path generation unit 17. In S08 and S09, the ECU 10 specifically performs Figure 5 The processing shown.

[0088] Figure 5 It shows Figure 3 This is a flowchart of an example of the processing of S08 and S09. Figure 5 The flowchart shown is used, for example, in autonomous driving Figure 1 ECU processing is executed.

[0089] like Figure 5 As shown, in S21, the ECU 10 determines, via the travel path generation unit 17, whether the vehicle's own path and the target path interfere with each other. If the ECU 10 determines that the vehicle's own path and the target path interfere (S21: Yes), the process proceeds to S22. If the ECU 10 determines that the vehicle's own path and the target path do not interfere (S21: No), the process proceeds to S25. In S21, for example, if multiple target paths are identified and at least one target path interferes with the vehicle's own path, the ECU 10 may determine that the vehicle's own path and the target path interfere.

[0090] In S22, the ECU 10 obtains the interference position and the interference time through the travel path generation unit 17. In S23, the ECU 10 generates a first travel path for avoiding the vehicle's interference through the travel path generation unit 17. In S24, the ECU 10 controls the vehicle's travel according to the first travel path through the travel control unit 18 so that the vehicle stops before the interference position. After processing S24, the ECU 10 ends. Figure 5 processing, move to Figure 3 Processing and end Figure 3 processing.

[0091] On the other hand, in S25, the ECU 10 generates a second travel path along the own vehicle path P1a through the travel path generation unit 17. In S26, the ECU 10 controls the vehicle's travel according to the second travel path through the travel control unit 18, and causes the vehicle to travel along the own vehicle path P1a. After the processing of S26, the ECU 10 ends. Figure 5 processing, move to Figure 3 Processing and end Figure 3 processing.

[0092] As described above, the automated driving path generation device 20 of the first embodiment calculates a predicted acceleration of a moving object for each predicted path based on multiple predicted paths and the speed of the moving object. Based on the comparison result of the predicted acceleration with an acceleration threshold, a target path, which is a predicted path used in path generation, is determined. A path is generated based on the own vehicle path and the target path. By using the target path determined based on the comparison result of the predicted acceleration with the acceleration threshold, a path can be generated that takes into account the feasibility of the moving object's predicted path. This allows for more appropriate automated driving path generation compared to a case where the feasibility of the moving object's predicted path is not considered.

[0093] In the automated driving path generation device 20, the predicted acceleration calculation unit 15 calculates, for each predicted path, the predicted lateral acceleration generated by the mobile object moving along the predicted path based on the curvature radius of the predicted path and the speed of the mobile object. The target path determination unit 16 determines as the target path any predicted path whose maximum predicted lateral acceleration among the multiple predicted paths is less than a lateral acceleration threshold. This allows for more appropriate automated driving path generation by leveraging the low probability of achieving a predicted path where the maximum predicted lateral acceleration generated by the mobile object is greater than the lateral acceleration threshold.

[0094] According to the first embodiment, the automated driving device 100 calculates a predicted acceleration of a moving object for each predicted path based on multiple predicted paths and the speed of the moving object. Based on the comparison result of the predicted acceleration with an acceleration threshold, a target path, which is a predicted path used in generating a travel path, is determined from among the predicted paths. A travel path is generated based on the own vehicle path and the target path. Thus, by using the target path determined based on the comparison result of the predicted acceleration with the acceleration threshold, a travel path can be generated that takes into account the feasibility of the moving object's predicted path. Therefore, compared to a case where the feasibility of the moving object's predicted path is not considered, a more appropriate travel path for automated driving can be generated. Furthermore, according to the automated driving device 100, if interference occurs between the own vehicle path and the target path, a first travel path (interference-avoiding travel path) is generated to avoid interference, and the vehicle is decelerated according to the first travel path. Therefore, the vehicle can be stopped before the interference location. Thus, since the vehicle is decelerated based on the feasibility of the moving object's predicted path, the vehicle can be decelerated more appropriately compared to a case where the feasibility of the moving object's predicted path is not considered.

[0095] Furthermore, even when the vehicle requires sudden deceleration to avoid a collision with a moving object on a predicted path that interferes with the vehicle's path, predicted paths where the moving object is actually unable to travel are not considered target paths. Consequently, a vehicle path is not generated for such predicted paths that interfere with the vehicle's path but are unattainable, eliminating the need for unnecessary sudden deceleration. This prevents the vehicle from experiencing reduced driving efficiency or fuel economy due to unnecessary rapid deceleration.

[0096] [Second embodiment]

[0097] Next, an automatic driving route generating device 20A and an automatic driving device 100A according to a second embodiment will be described with reference to the drawings. Figure 6 This is a block diagram showing an automatic driving route generation device according to a second embodiment. Figure 6 The automated driving path generation device 20A and automated driving device 100A shown differ from the first embodiment in the following respects: predicted longitudinal acceleration is used instead of predicted lateral acceleration; and the target path is determined by taking into account traffic regulations and priorities. Components identical or corresponding to those in the first embodiment are denoted by the same reference numerals, and their descriptions are omitted.

[0098] like Figure 6 As shown, the automatic driving route generating device 20A and the automatic driving device 100A according to the second embodiment include a map database 4A instead of the map database 4 and an ECU 10A including a predicted acceleration calculating unit 15A and a target route specifying unit 16A.

[0099] Map database 4A is configured essentially the same as map database 4. Map database 4A also stores, as map information, priority information that defines the traffic regulations' priorities for multiple vehicles on a road. Examples of traffic regulations' priorities include priority for vehicles on roads that intersect at intersections, priority for vehicles on each road when another road merges with another road, and priority for vehicles in adjacent lanes.

[0100] Figure 7A This is a schematic plan view illustrating a predicted path of a merging vehicle at a Y-junction. Figure 7B is an example ratio Figure 7A A schematic top view of the predicted path of a merging car traveling quickly at a Y-shaped intersection. Figure 7A In FIG, as an example of a moving body, another vehicle V4 is shown as a merging vehicle that is about to merge with the road R5 on which the own vehicle V1 is traveling. Figure 7B In the figure, another vehicle V5 is shown as a comparison Figure 7A The merging car is a merging car traveling at a low speed.

[0101] The path calculation unit 14 functions in the same manner as in the first embodiment. Figure 7A and Figure 7B As shown, the own vehicle V1, other vehicles V4 and other vehicles V5 are about to arrive at the Y-shaped intersection RY. Figure 7A and Figure 7B In the example shown in FIG. 1 , the path calculation unit 14 calculates a vehicle path P1b as the vehicle path for automatic driving of the vehicle V1 based on the position of the vehicle V1 on the map and the map information. The vehicle path P1b is a straight path for the vehicle V1 from the road R5 to the road R6.

[0102] exist Figure 7A In the example, the path calculation unit 14 calculates the predicted path P4a and the predicted path P4b as the predicted paths of the other vehicle V4. The predicted path P4a is the merging path of the other vehicle V4 that will merge onto the road R6 after temporarily stopping at the stop line (temporary stop position) SL set on the road R7. The predicted path P4b is the merging path of the other vehicle V4 that will merge onto the road R6 without temporarily stopping at the stop line SL. Figure 7B In the example shown in FIG. 1 , the path calculation unit 14 calculates predicted paths P5a and P5b as predicted paths for another vehicle V5. Predicted path P5a is the merging path for another vehicle V5 that temporarily stops at the stop line SL on road R7 and then merges onto road R6. Predicted path P5b is the merging path for another vehicle V5 that does not temporarily stop at the stop line SL and instead merges onto road R6.

[0103] The predicted acceleration calculation unit 15A calculates the predicted longitudinal acceleration as the predicted acceleration. The predicted longitudinal acceleration is the longitudinal acceleration that the moving body is predicted to generate when the moving body accelerates or decelerates along the predicted path. The longitudinal direction refers to the direction in which the moving body is traveling.

[0104] The predicted acceleration calculation unit 15A calculates the predicted longitudinal acceleration generated by the mobile body moving along the predicted path for each predicted path based on the temporary stop position or deceleration position of the mobile body on the predicted path and the speed of the mobile body. The temporary stop position is a position where the mobile body is considered to be temporarily stopped according to traffic regulations or the general concept of traffic society. Examples of temporary stop positions include the position of a stop line, a predetermined position before an intersection, and the position of an entrance or exit of a facility along the road. The deceleration position is a position where the mobile body is considered to be slowed down according to traffic regulations or the general concept of traffic society. Examples of deceleration positions include a position where the speed limit is changed to a lower speed and a predetermined position before entering a curve from a straight road.

[0105] exist Figure 7A In the example shown in FIG. 1 , the predicted acceleration calculation unit 15A calculates, as the predicted longitudinal acceleration, a predicted value of the negative acceleration that would be incurred by the other vehicle V4 when the other vehicle V4 decelerates along the predicted path P4a and stops at the stop line SL, based on the position of the stop line SL on the predicted path P4a and the speed of the other vehicle V4. Furthermore, the predicted acceleration calculation unit 15A calculates, as the predicted longitudinal acceleration, a predicted value of the negative acceleration that would be incurred by the other vehicle V4 when the other vehicle V4 decelerates along the predicted path P4b (e.g., a position where the other vehicle V4 slows down when merging) based on the deceleration position on the predicted path P4b and the speed of the other vehicle V4.

[0106] exist Figure 7B In the example shown in FIG. 1 , the predicted acceleration calculation unit 15A calculates, as the predicted longitudinal acceleration, a predicted value of the negative acceleration that would be incurred by the other vehicle V5 when the other vehicle V5 decelerates while traveling along the predicted path P5a and stops at the stop line SL, based on the position of the stop line SL on the predicted path P5a and the speed of the other vehicle V5. Furthermore, the predicted acceleration calculation unit 15A calculates, as the predicted longitudinal acceleration, a predicted value of the negative acceleration that would be incurred by the other vehicle V5 when the other vehicle V5 decelerates while traveling along the predicted path P5b, based on the deceleration position on the predicted path P5b and the speed of the other vehicle V5.

[0107] The target route identification unit 16A identifies, as the target route, a predicted route in which the maximum value of the magnitude of the predicted longitudinal acceleration among the plurality of predicted routes is smaller than the longitudinal acceleration threshold value.

[0108] The longitudinal acceleration threshold is a threshold of predicted longitudinal acceleration used to determine whether the predicted path of a moving object is a path with a low probability of being physically achieved. The longitudinal acceleration threshold can also be a pre-set parameter or mapping. The longitudinal acceleration threshold can also be set based on the road conditions where deceleration is required. For example, the longitudinal acceleration threshold can be set to a higher threshold when the moving object is supposed to stop at a temporary stop position than when the moving object is supposed to decelerate at a deceleration position. Furthermore, the longitudinal acceleration threshold can also be set based on the type of moving object.

[0109] Specifically, the target path determination unit 16A may determine whether the maximum value of the predicted longitudinal acceleration is less than a longitudinal acceleration threshold. If the target path determination unit 16A determines that the maximum value of the predicted longitudinal acceleration is less than the longitudinal acceleration threshold, the predicted path is determined as the target path. If the target path determination unit 16A determines that the maximum value of the predicted longitudinal acceleration is greater than the longitudinal acceleration threshold, the predicted path is not determined as the target path. Furthermore, in the above determination by the target path determination unit 16A, a signed value may be used as the predicted longitudinal acceleration, and the longitudinal acceleration threshold may also use a correspondingly signed value.

[0110] Here, in Figure 7A and Figure 7B In the example, based on the relationship between the vehicle V1 and other vehicles V4 and V5 approaching the Y-junction RY, traffic regulations require that the other vehicles V4 and V5 should temporarily stop at the stop line SL. The other vehicle V5 is traveling at a lower speed than the other vehicle V4. The speed of the other vehicle V5 is such that it can temporarily stop at the stop line SL. However, the speed of the other vehicle V4 is so fast that it is difficult to temporarily stop at the stop line SL in reality. In this case, Figure 7A In this case, the other vehicle V4 is forced to travel along predicted path P4b, and the likelihood of achieving predicted path P4a is lower than that of predicted path P4b. Therefore, the target path determination unit 16A determines that the maximum value of the predicted longitudinal acceleration for predicted path P4b is less than the longitudinal acceleration threshold, and that the maximum value of the predicted longitudinal acceleration for predicted path P4a is greater than the longitudinal acceleration threshold. The target path determination unit 16A determines predicted path P4b as the target path. On the other hand, the target path determination unit 16A does not determine predicted path P4a as the target path.

[0111] The target route determination unit 16A may further determine a target route in which the priority of a mobile object among the multiple predicted routes is higher than the priority of the vehicle. Specifically, the target route determination unit 16A obtains the vehicle priority and the mobile object priority in the relationship between the vehicle and the mobile object as priority information based on the vehicle's position on the map, the mobile object's position on the map, and the map information. For example, the priority can be a number assigned to the vehicle and all mobile objects in sequence, starting from 1, based on the priority of traffic regulations at the intersection the vehicle and the mobile object will approach.

[0112] The target path determination unit 16A may also determine whether the priority of the mobile object is higher than that of the vehicle. If the target path determination unit 16A determines that the priority of the mobile object is higher than that of the vehicle, it determines the predicted path as the target path. If the target path determination unit 16A determines that the priority of the mobile object is lower than that of the vehicle, it does not determine the predicted path as the target path.

[0113] exist Figure 7B, the speed of the other vehicle V5 is a speed at which it can temporarily stop at the stop line SL. Therefore, the object path determination unit 16A determines that the maximum value of the predicted longitudinal acceleration of the predicted paths P5a and P5b is less than the longitudinal acceleration threshold. The object path determination unit 16A determines the predicted paths P5a and P5b as object paths. When the number of determined object paths (predicted paths) is 2 or more, the object path determination unit 16A further determines an object path whose priority of the mobile body among the multiple predicted paths is higher than the priority of the vehicle. The object path determination unit 16A may also determine the object path using priority information for the predicted path that enters the road where the vehicle is traveling or the predicted path that enters the intersection where the vehicle is about to arrive.

[0114] exist Figure 7B In the example, the priority of the host vehicle V1 is 1, and the priority of the other vehicle V5 is 2. In other words, the priority of the other vehicle V5 is lower than that of the host vehicle V1. The target path determination unit 16A determines that the priority of the other vehicle V5 is lower than that of the host vehicle V1. The target path determination unit 16A does not select the predicted path P5b, which is the predicted path for entering the Y-intersection RY that the host vehicle V1 is about to approach, as the target path. As a result, the target path determination unit 16A selects the remaining predicted path P5a as the target path.

[0115] Incidentally, when the number of the identified target paths (predicted paths) is one, the target path identification unit 16A may omit identification of the target paths using the priority information. Figure 7A In the example of , the predicted path is narrowed down to one by determining the target path using the predicted longitudinal acceleration, and thus determination of the target path using the priority information is omitted.

[0116] Next, the processing of the automatic driving route generating device 20A and the ECU 10A of the automatic driving device 100A will be described with reference to the drawings.

[0117] Reference Figure 8 The processing of the ECU 10A will be described. Figure 8 It shows Figure 6 Flowchart of an example of automatic driving processing by ECU 10A. Figure 8 The flowchart shown is executed, for example, during automated driving.

[0118] like Figure 8 As shown, the automatic driving process of ECU10A is basically the same as Figure 3 The automatic driving process of ECU10A is the same as that of ECU10A. Figure 3The automatic driving processing of the ECU 10 is different. Therefore, these processing will be described below.

[0119] In S35, the ECU 10A of the automatic driving path generation device 20A and the automatic driving device 100A calculates the predicted acceleration of each predicted path through the predicted acceleration calculation unit 15A. The predicted acceleration calculation unit 15A calculates the predicted acceleration generated by the moving body moving along the predicted path for each predicted path based on the multiple predicted paths and the speed of the moving body. In S36, the ECU 10A determines the target path through the target path determination unit 16A. The target path determination unit 16A determines the target path based on the comparison result of the predicted acceleration and the acceleration threshold. In S35 and S36, the ECU 10A specifically performs Figure 9 and Figure 10 The processing shown.

[0120] Figure 9 It shows Figure 8 This is a flowchart of an example of the processing of S35 and S36. Figure 9 The flowchart shown is used, for example, in autonomous driving Figure 8 In the case of multiple predicted paths, Figure 8 The process can also be repeatedly performed for each predicted path.

[0121] like Figure 9 As shown, the ECU 10A performs the process of S41 for each predicted path as the process of S35 described above. In S41, the ECU 10A uses the predicted acceleration calculation unit 15A to calculate the predicted longitudinal acceleration for each predicted path based on the temporary stopping position or deceleration position of the mobile body on the predicted path and the speed of the mobile body. For example, the predicted acceleration calculation unit 15A calculates the acceleration generated by other vehicles V4 and V5 as they decelerate to temporarily stop at the stop line SL on predicted paths P4a and P5a as the predicted longitudinal acceleration.

[0122] The ECU 10A performs steps S42 through S44 for each predicted path as the aforementioned step S36. In S42, the ECU 10A, through the target path identification unit 16A, determines whether the maximum value of the predicted longitudinal acceleration is less than the longitudinal acceleration threshold. If the ECU 10A determines that the maximum value of the predicted longitudinal acceleration is less than the longitudinal acceleration threshold (S42: Yes), the process proceeds to S43. If the ECU 10A determines that the maximum value of the predicted longitudinal acceleration is greater than the longitudinal acceleration threshold (S42: No), the process proceeds to S44.

[0123] In S43 , the ECU 10A determines the predicted route as the target route through the target route determination unit 16A. On the other hand, in S44 , the ECU 10A does not determine the predicted route as the target route through the target route determination unit 16A.

[0124] After the processing of S43 or S44, the ECU 10A ends Figure 9 If the number of the identified target paths is 2 or more, the ECU 10A moves to Figure 10 If the number of the identified target paths is 1, the ECU 10A omits the processing of S51. Figure 10 processing, move to Figure 8 S37 processing.

[0125] Figure 10 This is a flowchart showing an example of additional target path determination processing. Figure 10 The flowchart shown is used, for example, in autonomous driving Figure 8 The ECU process is executed as an additional process to the process of S36. Figure 10 The process may also be repeatedly performed on each of a plurality of prediction paths.

[0126] like Figure 10 As shown, the ECU 10A can perform steps S51 to S54 for each predicted route as an additional step to S36. In S51, the ECU 10A obtains priority information via the target route determination unit 16A. The target route determination unit 16A obtains the vehicle's priority and the mobile object's priority in the relationship between the vehicle and the mobile object as priority information based on, for example, the vehicle's position on a map, the mobile object's position on a map, and the map information.

[0127] In S52, the ECU 10A determines, via the target route identification unit 16A, whether the priority of the mobile object is higher than that of the vehicle. If the ECU 10A determines that the priority of the mobile object is higher than that of the vehicle (S52: Yes), the process proceeds to S53. If the ECU 10A determines that the priority of the mobile object is lower than that of the vehicle (S52: No), the process proceeds to S54.

[0128] In S53, the ECU 10A determines the predicted path as the target path through the target path determination unit 16A. On the other hand, in S54, the ECU 10A does not determine the predicted path as the target path through the target path determination unit 16A. After the processing of S53 or S54, the ECU 10A ends. Figure 10 processing and move to Figure 8 S37 processing.

[0129] As described above, in the second embodiment of the automated driving path generation device 20A, the predicted acceleration calculation unit 15A calculates, for each predicted path, the predicted longitudinal acceleration generated by the mobile body moving along the predicted path based on the temporary stop position or deceleration position of the mobile body on the predicted path and the speed of the mobile body. The target path determination unit 16A determines as the target path the predicted path whose maximum value of the predicted longitudinal acceleration among the multiple predicted paths is less than the longitudinal acceleration threshold. This allows for more appropriate generation of automated driving paths by leveraging the fact that predicted paths where the maximum value of the predicted longitudinal acceleration generated by the mobile body is greater than the longitudinal acceleration threshold are considered to have a low probability of being achieved.

[0130] In the automated driving route generation device 20A, map information includes priority information that determines the order of priority for multiple vehicles under traffic regulations. The target route determination unit 16A further determines a target route in which the priority of a mobile object among the multiple predicted routes is higher than the priority of a vehicle. This allows for more appropriate automated driving route generation by leveraging the fact that the predicted route, which is considered to have a lower probability of being realized, is lower than the priority of a vehicle.

[0131] [Modification]

[0132] While preferred embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments. The present invention can be implemented in various forms represented by the above-described embodiments and variously modified and improved based on the knowledge of those skilled in the art.

[0133] For example, the target path determination unit 16 or 16A may further determine a target path, among the multiple predicted paths, that extends in the direction determined by the direction indication display. In this case, the vehicle exterior environment identified by the vehicle exterior environment recognition unit 12 may also include a direction indication display of a moving object. A direction indication display is a display used to indicate the direction of the moving object's planned travel to the surroundings. The vehicle exterior environment recognition unit 12 may also identify a direction indication display based on an image captured by a camera that includes the moving object. An example of a direction indication display is a flashing turn signal indicator of another vehicle. Alternatively, a hand signal display may be used.

[0134] For example, the ECU 10 or 10A may also be Figure 4 Processing, Figure 9 Processing or Figure 10 If the number of target paths determined in the process is 2 or more, move to Figure 11 The S61 process. Figure 11 This is a flowchart showing another example of the additional object path determination process. Figure 11The process may also be repeatedly performed on each of a plurality of prediction paths.

[0135] like Figure 11 As shown, in S61, the ECU 10, 10A acquires direction indication information via the vehicle exterior environment recognition unit 12. The vehicle exterior environment recognition unit 12 acquires direction indication information based on, for example, an image captured by a camera including a moving object.

[0136] In S62, the ECU 10, 10A determines via the target route identification unit 16, 16A whether the predicted route extends in the direction specified by the direction indication display. If the ECU 10, 10A determines that the predicted route extends in the direction specified by the direction indication display (S62: Yes), the process proceeds to S63. If the ECU 10, 10A determines that the predicted route does not extend in the direction specified by the direction indication display (S62: No), the process proceeds to S64.

[0137] In S63, the ECU 10, 10A determines the predicted path as the target path through the target path determination unit 16, 16A. On the other hand, in S64, the ECU 10, 10A does not determine the predicted path as the target path through the target path determination unit 16, 16A. After the processing of S63 or S64, the ECU 10, 10A ends. Figure 11 processing, move to Figure 3 S07 processing or Figure 8 S37 processing.

[0138] In this manner, the vehicle exterior environment recognition unit 12 recognizes the direction indication of the moving object, and the target path determination unit 16, 16A further determines, as the target path, a predicted path extending in the direction indicated by the direction indication from among the plurality of predicted paths. This allows for more appropriate generation of an autonomous driving path by taking advantage of the fact that predicted paths extending in directions not indicated by the direction indication are considered less likely to be achieved.

[0139] In addition, for example, the object path determination unit 16, 16A may further determine an object path extending in a direction corresponding to the offset of the lateral position among the multiple predicted paths. In this case, the vehicle exterior environment identified by the vehicle exterior environment recognition unit 12 may also include the offset of the lateral position of the mobile body. The lateral position refers to the position of the mobile body in the road width direction of the road on which the mobile body is traveling. In the case where, for example, other vehicles traveling in opposite lanes are assumed to be mobile bodies, the lateral position may also be the position of the vehicle in the lane width direction of the opposite lane on which the other vehicles are traveling. The lane width direction refers to the direction on the road surface that is orthogonal to the white lines forming the lanes of the road. The offset of the lateral position refers to the offset of the lateral position of the mobile body relative to the entire road width direction of the road on which the mobile body is traveling. The vehicle exterior environment recognition unit 12 may also identify the offset of the lateral position based on which of a pair of end portions in the road width direction of the road on which the mobile body is traveling is close. When assuming that there is another vehicle traveling in the opposite lane as a moving object, the fact that the lateral position of the moving object in the opposite lane is close to the center line can predict that the moving object is likely to move in a manner that crosses over to the lane in which the vehicle is traveling or exceeds the lane in which the vehicle is traveling.

[0140] For example, the ECU 10 or 10A may also be Figure 4 Processing, Figure 9 Processing, Figure 10 Processing or Figure 11 If the number of target paths determined in the process is 2 or more, move to Figure 12 The S71 process. Figure 12 This is a flowchart showing another example of the additional object path determination process. Figure 12 The process may also be repeatedly performed on each of a plurality of prediction paths.

[0141] like Figure 12 As shown, in S71, the ECU 10, 10A recognizes the lateral position shift via the vehicle exterior environment recognition unit 12. The vehicle exterior environment recognition unit 12 recognizes the lateral position shift based on, for example, an image captured by a camera including the moving object.

[0142] In S72, the ECU 10, 10A determines, via the target path identification unit 16, 16A, whether the predicted path extends in the direction corresponding to the lateral position shift. If the ECU 10, 10A determines that the predicted path extends in the direction corresponding to the lateral position shift (S72: Yes), the process proceeds to S73. If the ECU 10, 10A determines that the predicted path does not extend in the direction corresponding to the lateral position shift (S72: No), the process proceeds to S74.

[0143] In S73, the ECU 10, 10A determines the predicted path as the target path through the target path determination unit 16, 16A. On the other hand, in S74, the ECU 10, 10A does not determine the predicted path as the target path through the target path determination unit 16, 16A. After the processing of S73 or S74, the ECU 10, 10A ends. Figure 12 processing, move to Figure 3 S07 processing or Figure 8 S37 processing.

[0144] In this manner, the vehicle exterior environment recognition unit 12 recognizes a lateral position shift in a direction intersecting the vehicle's travel direction, and the target path determination units 16 and 16A further determine a target path, from among the plurality of predicted paths, that extends in a direction corresponding to the lateral position shift. This allows for more appropriate generation of an autonomous driving path by utilizing the fact that predicted paths extending in directions not corresponding to the lateral position shift are considered less likely to be achieved.

[0145] In addition, for example, the object path determination unit 16, 16A may further determine the object path based on the recognition result of the lighting state of the signal lights in multiple predicted paths. In this case, the vehicle exterior environment recognized by the vehicle exterior environment recognition unit 12 may also include the recognition result of the lighting state of the signal lights. The vehicle exterior environment recognition unit 12 may also recognize the lighting state of the signal lights in front of the vehicle (whether it is a lighting state that allows passing or a lighting state that prohibits passing, etc.) based on the image of the camera of the external sensor 2. The vehicle exterior environment recognition unit 12 may also recognize the lighting state of the signal lights based on infrastructure information that can be obtained via the communication unit. The object path determination unit 16, 16A may also determine whether the predicted path extends in the direction in which the lighting state of the signal lights is a lighting state in which passing is possible. The object path determination unit 16, 16A may also determine the predicted path as the object path when it is determined that the predicted path extends in the direction in which the lighting state of the signal lights is a lighting state in which passing is possible. On the other hand, the target path identification unit 16 or 16A may not identify the predicted path as the target path when determining that the predicted path extends in a direction in which the lighting state of the traffic light is the lighting state prohibiting passage.

[0146] In this manner, the vehicle exterior environment recognition unit 12 recognizes the lighting state of the traffic light, and the target path determination unit 16, 16A further determines, as the target path, a predicted path extending in the direction where the traffic light's lighting state is in a passable state from among the multiple predicted paths. This allows for more appropriate generation of an autonomous driving path by utilizing the fact that the predicted path extending in the direction where the traffic light's lighting state is in a passable state is considered to have a low probability of being realized.

[0147] Alternatively, for example, the target path determination unit 16 or 16A may further determine the target path based on whether the road on which the vehicle is located is one-way traffic among the multiple predicted paths. If the predicted path of the mobile object extends along the road on which the vehicle is located, the mobile object is traveling in the wrong direction of one-way traffic, and the likelihood of achieving the predicted path is considered low. Therefore, the target path determination unit 16 or 16A may not determine the predicted path as the target path.

[0148] In this manner, the vehicle exterior environment recognition unit 12 recognizes the lighting state of the traffic light, and the target path determination unit 16, 16A further determines, as the target path, a predicted path extending in the direction where the traffic light's lighting state is in a passable state from among the multiple predicted paths. This allows for more appropriate generation of an autonomous driving path by utilizing the fact that the predicted path extending in the direction where the traffic light's lighting state is in a passable state is considered to have a low probability of being realized.

[0149] In the above embodiment, the first travel path is a travel path that stops the vehicle before the interference position, but the present invention is not limited to this. For example, if the own vehicle path interferes with the target path, the travel path generation unit 17 may generate the first travel path according to a steering plan that moves the vehicle traveling along the own vehicle path away from the moving object based on the vehicle speed, the own vehicle path, and the interference position. Furthermore, the first travel path may include both a speed plan that stops the vehicle before the interference position and a steering plan that moves the vehicle away from the moving object.

[0150] In the above embodiment, the conditions for determining whether a path is a target path include, for example, the comparison result of the predicted lateral acceleration with the lateral acceleration threshold, or the comparison result of the predicted longitudinal acceleration with the longitudinal acceleration threshold. However, the present invention is not limited to these conditions. For example, the target path determination unit may determine a path as a target path based on the comparison result of the combined predicted acceleration calculated from the predicted lateral acceleration and the predicted longitudinal acceleration with the acceleration threshold for the combined predicted acceleration.

[0151] In the above embodiment, the predicted lateral acceleration and the predicted longitudinal acceleration are exemplified as the predicted acceleration, but the present invention is not limited thereto. For example, a predicted value of acceleration in the vehicle height direction may be used as the predicted acceleration.

[0152] In the second embodiment, the target route is determined using the priority order based on traffic regulations, but this can be omitted. The first embodiment and the second embodiment can also be combined with each other. The above-mentioned modified examples can also be appropriately combined to determine the target route.

[0153] In the first embodiment described above, the maximum value of the predicted lateral acceleration is compared with the lateral acceleration threshold, but the present invention is not limited to this. For example, in the determination of the object path determination unit, a value with a positive or negative sign may be used as the predicted lateral acceleration, and the lateral acceleration threshold may also use a threshold with a corresponding positive or negative sign. In addition, the maximum value does not have to be used. In the second embodiment described above, the maximum value of the predicted longitudinal acceleration is compared with the longitudinal acceleration threshold, but the present invention is not limited to this. For example, in the determination of the object path determination unit, a value with a positive or negative sign may be used as the predicted longitudinal acceleration, and the longitudinal acceleration threshold may also use a threshold with a corresponding positive or negative sign. In addition, the maximum value does not have to be used. In short, the object path determination unit can determine the object path based on the comparison result of the predicted acceleration and the acceleration threshold.

[0154] While the above embodiment illustrates an autonomous driving system equipped with an autonomous driving path generator, the present invention is not limited thereto. The present invention can also be implemented in a configuration that omits at least the actuator 5 and the travel control unit 18, providing only the functionality of the autonomous driving path generator. In this case, for example, the path generated by the autonomous driving path generator can be transmitted from the autonomous driving path generator to the autonomous driving vehicle via a communication network, etc., and used to enable the autonomous driving vehicle to perform autonomous driving.

Claims

1. An automatic driving path generation device for generating an automatic driving path for a vehicle. The automatic driving path generation device comprises: a moving object recognition unit configured to recognize a moving object located around the vehicle; a path calculation unit that calculates a self-vehicle path for automatic driving of the vehicle and a plurality of predicted paths for the mobile object based on the position of the vehicle on the map, the position of the mobile object on the map, and map information; a predicted acceleration calculation unit that calculates, for each of the plurality of predicted paths and a speed of the mobile object, a predicted acceleration generated by the mobile object moving along the predicted path; a target path determination unit that determines, based on a comparison result between the predicted acceleration and an acceleration threshold value, a target path, which is the predicted path used in generating the travel path, among the plurality of predicted paths; as well as a travel path generating unit that generates the travel path based on the own vehicle path and the target path, The moving object recognition unit recognizes a deviation of the lateral position of the moving object in a direction intersecting the traveling direction, wherein the lateral position deviation refers to a deviation of the lateral position of the moving object relative to the entire width direction of the road on which the moving object is traveling. The target path determination unit further determines, as the target path, the predicted path extending in a direction corresponding to the shift in the lateral position among the plurality of predicted paths.

2. The automatic driving route generation device according to claim 1, The predicted acceleration calculation unit calculates, for each predicted path, a predicted lateral acceleration generated by the mobile body moving along the predicted path based on the curvature radius of the predicted path and the speed of the mobile body. The target path determination unit determines, as the target path, a predicted path in which the maximum value of the magnitude of the predicted lateral acceleration among the plurality of predicted paths is smaller than a lateral acceleration threshold value.

3. The automatic driving route generation device according to claim 1 or 2, The predicted acceleration calculation unit calculates, for each predicted path, a predicted longitudinal acceleration of the moving body moving along the predicted path based on a temporarily stopped position or a decelerated position of the moving body on the predicted path and the speed of the moving body. The target path determination unit determines, as the target path, a predicted path in which the maximum value of the magnitude of the predicted longitudinal acceleration among the plurality of predicted paths is smaller than a longitudinal acceleration threshold value.

4. The automatic driving route generation device according to claim 1 or 2, The map information includes priority information for determining priorities of a plurality of vehicles under traffic regulations. The target path determination unit further determines, as the target path, a predicted path in which the priority order of the mobile body among the plurality of predicted paths is higher than the priority order of the vehicle.

5. The automatic driving route generation device according to claim 1 or 2, The moving object recognition unit recognizes the direction indication display of the moving object, The target path determination unit further determines, as the target path, the predicted path extending in the direction determined by the direction indication display among the plurality of predicted paths.

6. An automatic driving device that enables a vehicle to automatically drive according to a generated travel path, the automatic driving device comprising: a moving object recognition unit configured to recognize a moving object located around the vehicle; a path calculation unit that calculates a self-vehicle path for automatic driving of the vehicle and a plurality of predicted paths for the mobile object based on the position of the vehicle on the map, the position of the mobile object on the map, and map information; a predicted acceleration calculation unit that calculates, for each of the plurality of predicted paths and a speed of the mobile object, a predicted acceleration generated by the mobile object moving along the predicted path; a target path determination unit that determines, based on a comparison result between the predicted acceleration and an acceleration threshold value, a target path, which is the predicted path used in generating the travel path, among the plurality of predicted paths; a travel path generating unit configured to generate the travel path based on the own vehicle path and the target path; as well as a driving control unit configured to cause the vehicle to automatically drive according to the travel path; The moving object recognition unit recognizes a deviation of the lateral position of the moving object in a direction intersecting the traveling direction, wherein the lateral position deviation refers to a deviation of the lateral position of the moving object relative to the entire width direction of the road on which the moving object is traveling. The target path determination unit further determines, as the target path, the predicted path extending in a direction corresponding to the offset of the lateral position among the plurality of predicted paths. The travel path generating unit, When the own vehicle path interferes with the target path, calculating the interference position between the own vehicle path and the target path, generating an interference-avoiding travel path as the travel path such that the vehicle stops before the interference position based on the vehicle speed, the own vehicle path, and the interference position; The travel control unit decelerates the vehicle according to the interference avoidance travel path.

Images