Vehicle control device

Abstract

A processor (control system calculation unit 108) of a vehicle control device acquires (approach risk calculation unit 109) the degree of influence (contact influence degree) that is exerted on a host vehicle due to contact between the host vehicle and a boundary structure, constituting a boundary of a region where the host vehicle travels, and that corresponds to the kind of the boundary structure. The processor acquires (approach risk calculation unit 109) an approach risk that indicates a risk occurring when the host vehicle approaches the boundary structure and that corresponds to the combination of the height of the boundary structure and the degree of influence. The processor changes (travel route calculation unit 110) an approachable distance (margin) of the host vehicle to the boundary structure on the basis of the approach risk, or changes (notification control unit 112) a notification content to a driver of the host vehicle.

Description

Vehicle control device

[0001] The present invention relates to a vehicle control device.

[0002] Patent No. 7230190 (Patent Document 1) describes a technology that provides assistance for passing obstacles such as oncoming vehicles and parked vehicles on narrow roads that are separated on both sides by boundary structures such as walls, curbs, and guardrails.

[0003] In Patent Document 1, three-dimensional position information of boundary components and obstacles is acquired, and passability is determined taking into consideration the height position as well. In addition, if the obstacle is determined to be a moving object or a person or vehicle, the clearance area in the vehicle width direction is changed so as to move away from the obstacle.

[0004] This allows for passing assistance when the vehicle is traveling on a narrow road, which provides the driver with peace of mind while also helping to avoid collisions and wheel run-off.

[0005] Patent No. 7230190

[0006] Patent Document 1 assumes that the robot will not come into contact with boundary structures and obstacles, so while it can move away from objects that pose a high risk if approached, it cannot achieve operations such as moving closer to objects that would have a small impact if contacted.

[0007] However, in an actual road environment, there are road boundaries, such as grassy areas or hedges, that have little impact if the vehicle comes into contact with them, and road boundaries, such as low steps or unpaved shoulders, that the vehicle can run over. Even when coming into contact with the same type of boundary component, tires and wheels are less likely to be damaged than, for example, painted car bodies or easily damaged door mirrors. Thus, the impact varies greatly depending on which part of the vehicle comes into contact with the boundary component.

[0008] Therefore, assuming that the driver will not come into contact with boundary structures, a narrow road that the driver would otherwise judge to be passable may be determined to be impassable. Even if the road is determined to be passable, the driver may pass through with an excessively large gap between the road and the boundary structure. In this way, if a route that differs from the route the driver envisions is calculated, it can cause the driver to feel uncomfortable or distrustful.

[0009] In view of the above-mentioned problems, an object of the present invention is to provide a vehicle control device that can improve passing assistance on narrow roads.

[0010] In order to achieve the above-mentioned object, the vehicle control device of the present invention is equipped with a processor that acquires the degree of impact on the host vehicle due to contact between the host vehicle and a boundary structure that constitutes the boundary of the area in which the host vehicle is traveling, corresponding to the type of boundary structure that constitutes the boundary of the area in which the host vehicle is traveling, acquires an approach risk that indicates the risk when the host vehicle approaches the boundary structure, corresponding to a combination of the height of the boundary structure and the degree of impact, and changes the approachable distance of the host vehicle from the boundary structure or changes the content of the notification to the driver of the host vehicle based on the approach risk.

[0011] According to the present invention, there is provided a vehicle control device that can improve passing assistance on narrow roads. Problems, configurations, and effects other than those described above will become apparent from the following description of the embodiments.

[0012] 1 is a block diagram showing a schematic configuration of a vehicle control device of the present embodiment. FIG. 1 is a bird's-eye view showing an example of a situation in which a vehicle equipped with the vehicle control device shown in FIG. 1 travels through a narrow road. FIG. 2 is a front view showing an example of the situation in FIG. 2. FIG. 1 is a bird's-eye view showing another example of a situation in which a vehicle equipped with the vehicle control device shown in FIG. 1 travels through a narrow road. FIG. 4 is a front view showing an example of the situation in FIG. 4. FIG. 5 is a flowchart showing the flow of processing by a program executed by a control system calculation unit. FIG. 6 is a table showing an example of an approach risk and a host vehicle deceleration ratio according to boundary height and contact influence degree. FIG. 7 is a table showing an example of a contact influence degree according to the type of road boundary. FIG. 7 is a bird's-eye view showing an example of a method for calculating a driving path when passing a parked vehicle according to the present embodiment. FIG. 8 is a table showing an example of a margin for a road boundary according to an approach risk. FIG. 8 is a bird's-eye view showing another example of a method for calculating a driving path when passing a parked vehicle according to the present embodiment. FIG. 9 is a diagram showing an example of the driving speed of the host vehicle according to the difference between road width and vehicle width and the deceleration ratio. FIG. 10 is a diagram for explaining a means for determining whether or not a road width is wide enough for passing. FIG. 11 is a table showing an example of an alarm according to the approach risk of a road boundary and distance.

[0013] An embodiment of the present invention will now be described, which relates to a narrow road passing assistance function and method for a vehicle.

[0014] 1 is a schematic diagram of the configuration of a vehicle control device mounted on a vehicle. A recognition system calculation unit 105 acquires external environment recognition information from an external environment recognition unit 101, self-position information on a map from a self-position recognition unit 102, and map information around the vehicle from a high-precision map 103.

[0015] Here, the external environment recognition unit 101 is configured with one or more of a stereo camera, a monocular camera, LiDAR (Laser Imaging Detection and Ranging), millimeter-wave radar, and ultrasonic radar, and acquires external environment recognition information for detecting objects present around the vehicle. For reasons described below, a configuration using a stereo camera is preferable. The self-location recognition unit 102 is configured with a Global Navigation Satellite System (GNSS) or an inertial navigation system, and acquires the vehicle's own position on a map. The self-location recognition unit 102 may be configured to use a technique such as so-called map matching, which acquires the vehicle's position on a map by comparing the external environment recognition information acquired by the external environment recognition unit 101 with map information registered in a high-precision map 103. The high-precision map 103 is map information in which road environment information is recorded, and includes at least information on road shape, the number of lanes, and lane width, and may also include information on the location and type of road boundaries.

[0016] The recognition system calculation unit 105 has an obstacle detection unit 106 and a road boundary identification unit 107. The obstacle detection unit 106 detects obstacles present at least ahead of the vehicle based on the external environment recognition information acquired from the external environment recognition unit 101, and acquires the position, attitude angle, speed, angular velocity, three-dimensional shape, and type of the obstacle. Obstacles to be detected include, for example, oncoming vehicles, parked vehicles, pedestrians, bicycles, fallen objects, and other objects that are not fixed to the road.

[0017] The road boundary identification unit 107 identifies road boundaries that exist in the road width direction and separate the road from road shoulders, sidewalks, off-road property, etc. Road boundaries include road structures installed to indicate road boundaries, such as dividing lines, guardrails, sidewalks, road shoulders, and bollards (structures such as fences and poles that prevent vehicles from entering), as well as structures fixed to the road surface, such as hedges, vegetation, building walls, fences, utility poles, convex mirrors, and signs. The road boundary identification unit 107 acquires the position, height, and type of these road boundaries in the road width direction.

[0018] In this way, the recognition system calculation unit 105 measures the three-dimensional positions and shapes of obstacles and road boundaries and identifies their types. The position and shape of an object are measured to prevent contact between the vehicle and the object, but when measuring an object that has a shape that is partially convex, it may be impossible to detect the convex part unless a sensor with high resolution is used for measurement.

[0019] The type identification is performed to determine whether the detected road boundary is something that will cause damage to the vehicle body when it comes into contact with it, such as a curb or wall, or something that will not damage the vehicle when it comes into contact with it, such as grass or a hedge.

[0020] Therefore, it is desirable that the sensor constituting the external environment recognition unit 101 includes a stereo camera that has both high resolution and object identification capability. In this configuration, for example, the parallax between the left and right cameras constituting the stereo camera is used to measure the three-dimensional position and shape of the detected object, and the object is identified by an image recognition technology such as deep learning using the obtained image information.

[0021] In addition, the configuration may be a combination of LiDAR, which can measure the three-dimensional position and shape of a detected object with high resolution, and a monocular camera, which can identify objects using image recognition technology, or a configuration may be a combination of a stereo camera and LiDAR, which complement each other's obtained information.

[0022] Furthermore, instead of the road boundary identification unit, a configuration may be adopted in which the shape and type of road boundaries are acquired from boundary information included in the map information based on the self-location information acquired by the self-location recognition unit 102 and the map information acquired from the high-precision map 103. In this case, the external environment recognition information acquired by the external environment recognition unit 101 and the boundary information included in the map information are compared, and information determined to be the same object may be used in combination.

[0023] From the vehicle sensor 104, information on the host vehicle such as the traveling speed, yaw rate, steering angle, etc. of the host vehicle is acquired.

[0024] In the control system calculation unit 108, the approach risk calculation unit 109 calculates the risk of the host vehicle approaching an obstacle and a road boundary, and the driving path calculation unit 110 calculates a driving path for realizing passing. Then, the operation amount calculation unit 111 calculates the operation amounts of steering, driving, and braking required to realize the driving path, and these are realized by the actuator 113. Furthermore, the notification control unit 112 determines the notification content according to the approach risk and notifies the driver using the driver notification device 114. Here, the driver notification device 114 can use, for example, a method of displaying the information on a display or the like, or a method of notifying the driver using sound, vibration, light, etc.

[0025] The recognition system calculation unit 105 and the control system calculation unit 108 are configured by, for example, a processor such as a CPU (Central Processing Unit), a storage device such as a memory, and a communication device such as an input / output circuit. The functions of the control system calculation unit 108 and the recognition system calculation unit 105 may be realized by separate processors or by a single processor.

[0026] 2 shows an example of a scene to which this embodiment is applied, in which a vehicle 10 traveling on a narrow road passes an oncoming vehicle 11. FIG. 3 shows a front view of the same scene as seen from the left side of the paper in FIG. 2.

[0027] In this scene, there are walls 21 and 22 on the left and right that serve as boundaries. In addition, there is a low step 23 on the left side of the vehicle 10 in the direction of travel, and although this step is higher than the road 20, it is low enough relative to the tire radius of the vehicle, so it is possible for the vehicle 10 to climb up it. In this case, the road boundary on the right side of the vehicle 10 in the direction of travel is the wall 22 and the oncoming vehicle 11, and the road boundary on the left side of the vehicle in the direction of travel is the wall 21 and the low step 23. In this way, multiple road boundaries may exist even in a single scene.

[0028] Regarding the boundary on the right side of the traveling direction of the host vehicle 10, it is clear that the closest of the oncoming vehicle 11 and the wall 22 is the boundary because an accident will occur if the oncoming vehicle 11 comes into contact with either of them. On the other hand, regarding the boundary on the left side of the traveling direction, as shown in Figure 3, the distance in the road width direction between the oncoming vehicle 11 and the low step 23 is greater than the width of the host vehicle 10, so it is possible for the oncoming vehicle 11 to pass the low step 23, but it is also possible for the oncoming vehicle 11 to pass the low step 23 while climbing over the low step 23.

[0029] 4 and 5 show the same road environment as the scenes shown in Figures 2 and 3, but an oncoming vehicle 11 is traveling near the center of the roadway 20. As shown in Figure 5, in this scene, the distance in the road width direction between the oncoming vehicle 11 and the low step 23 is smaller than the width of the host vehicle 10, so the host vehicle cannot pass the oncoming vehicle 11 by traveling between the low step 23. Therefore, the host vehicle must climb up onto the low step 23 to pass the oncoming vehicle 11.

[0030] As described above, even in the same road environment, there are situations where the routes for passing each other are different and multiple routes exist. Therefore, hereinafter, the processing content for realizing passing by the control system calculation unit 108 will be explained using the flowchart shown in FIG.

[0031] In S201, the recognition system calculation unit 105 detects an obstacle on the path of the host vehicle using the method described above. The obstacle detected in the scenes shown in Figures 2 to 5 is an oncoming vehicle 11. In Figure 4, the door mirrors of the host vehicle 10 and the oncoming vehicle 11 are the same height, so the door mirrors touch each other. However, if the door mirrors of the host vehicle 10 and the oncoming vehicle 11 are different heights, the door mirrors do not touch each other, and the host vehicle 10 can get closer to the oncoming vehicle 11. Therefore, rather than simply representing the oncoming vehicle 11 as a rectangle that contains it, it is desirable to be able to distinguish between the vehicle body and protruding parts such as the door mirrors. Note that even if the oncoming vehicle 11 is represented in a simplified form that contains it within a rectangle, the present invention is still applicable as long as a conservative path is obtained. Furthermore, depending on the shape of the obstacle, a model other than a rectangle, such as a cylindrical model, may be used.

[0032] In step S202, the position, height, and type of road boundaries along the road width direction of the vehicle are identified using the method described above. In the scenes shown in Figures 2 to 5, the road boundaries identified are walls 21 and 22, a small step 23, and an oncoming vehicle 11 detected as an obstacle.

[0033] In S203, the approach risk is evaluated based on the height and type of the boundary, based on the tables shown in FIGS.

[0034] Figure 7 shows the risk of approach depending on the height of the road boundary and the impact of contact, and the higher the risk of approach, the further away from the boundary the route calculated will be. Because the approach to contact changes depending on the boundary height, the risk is set in four stages: zero to bottom edge of wheel, bottom edge of wheel to bottom edge of body, and bottom edge of body to bottom edge of mirror.

[0035] For example, if the height of the boundary is lower than the bottom edge of the wheel, it is the side of the tire that comes into contact with the boundary, and there is little chance of scratches. On the other hand, if the height of the boundary is higher than the bottom edge of the body, the painted body or easily damaged door mirrors will come into contact, so the impact of contact will be significant. Also, if the height of the boundary is between the bottom edge of the wheel and the bottom edge of the body, it is the wheel that comes into contact with the boundary, and there is less chance of scratches compared to the body or door mirrors. As the degree of impact differs depending on the part that comes into contact with the boundary, the approach risk is set according to the height.

[0036] Furthermore, since the degree of impact of contact varies depending on the type of road boundary, the degree of impact that the vehicle 10 will receive if it contacts the road boundary is set according to the type of road boundary identified, as shown in Figure 8. For example, the degree of impact is set to small for low bumps that can be driven over or grass, which will not damage the vehicle body even if the vehicle contacts the road boundary. Furthermore, the degree of impact is set to large for guardrails and utility poles, which are likely to cause damage to the vehicle body even with slight contact. A bump or hedge of approximately 15 cm may damage the vehicle body in some cases, so the degree of impact is set to medium. Markings and road shoulders are set to separate driving areas on the road, but crossing them does not affect the vehicle body, so the degree of impact is set to zero.

[0037] 7 and 8 are merely examples, and may be changed depending on the shape and material of the vehicle body, the designer's ideas, etc. Also, the settings may be changed according to the driver's intentions.

[0038] Furthermore, instead of determining the contact impact degree using the results of identifying the road boundary type in S202 and using Fig. 8, the contact impact degree may be directly acquired by image recognition. In this case, the contact impact degree may be learned when learning data for image recognition is learned, and the contact impact degree may be directly output during image recognition.

[0039] In S204, a combination of boundaries through which the host vehicle can proceed is selected. For example, in the scenes shown in Figures 2 and 3, the road is wide enough for the host vehicle to pass the oncoming vehicle 11 and the small step 23, so the step 23 is selected as the left boundary and the oncoming vehicle 11 as the right boundary. On the other hand, in the scenes shown in Figures 4 and 5, there is not enough width between the oncoming vehicle 11 and the small step 23 for the host vehicle to pass each other, but the small step 23 can be climbed up, and there is a wall 21 further outward. Therefore, the wall 21 is selected as the left boundary and the oncoming vehicle 11 is selected as the right boundary.

[0040] 4 and 5, if there is no boundary object such as a wall 21 outside the small step 23, then either the left boundary is null and only the right boundary is considered, or a virtual boundary with an approach risk of "1: no risk" is set at a certain distance away. Also, if no combination of boundaries provides a road wide enough for the vehicle to pass each other, then the combination that provides the widest road width is selected.

[0041] In S205, it is determined whether the road width between the combination of boundaries selected in S204 is greater than the vehicle width of the vehicle 10 plus the margins for the left and right boundaries described below, i.e., whether the road width is wide enough for two vehicles to pass each other. If the road width is wide enough for two vehicles to pass each other, S206 to S208 are executed in order.

[0042] In S206, a driving route is calculated based on the approach risk evaluated in S203. The method for calculating the driving route will be explained using FIG. 9. FIG. 9 shows a scene in which the host vehicle 10 passes between a parked vehicle 12 with an approach risk of "5: Contact is not permitted" and a hedge 24 with an approach risk of "3: Contact should be avoided." The left and right boundaries in this case are 25L and 25R (shown by dotted lines in FIG. 9). 26L and 26R are boundaries offset from the left and right boundaries 25L and 25R by margins w1L and w1R, respectively, which are set based on the approach risk of each boundary from the table in FIG. 10. Then, a driving route 27c (shown by a dashed line in FIG. 9) is set so that the host vehicle travels through the center between w1L and w1R. This allows a route to be calculated that is at least the margin away from the left and right boundaries.

[0043] The transition curve 27b is used to interpolate between the current travel path 27a of the vehicle 10 and the travel path 27c for passing each other. To facilitate path tracking, the transition curve 27b is given so that it changes continuously at least up to the curvature using a polynomial function, a clothoid curve, or the like.

[0044] The margin shown in Figure 10 is set so that the margin increases as the risk of approach increases. This reduces the possibility of accidental contact when deviating from the calculated driving route, and also makes it possible to calculate a route that is in line with the driver's intention to stay away from objects that could damage the vehicle body if they come into contact.

[0045] 11, a route may be calculated that is offset from the boundary closer to the host vehicle (the right boundary 25R in FIG. 11) by the sum of a margin w1R and half the vehicle width w10 of the host vehicle 10. In this case, too, the margin is taken into account in determining whether or not passage is possible, so a route that is at least the margin away from the left boundary 25L is always calculated.

[0046] When the example shown in Figure 9 is used, a route is calculated that is as far away as possible from both the left and right boundaries, whereas when the example shown in Figure 11 is used, a route is calculated that minimizes the amount of movement of the vehicle 10 in the road width direction.

[0047] In addition, when the object to be passed is moving, the transition curve 27b may be set so that the vehicle reaches a lateral position on the travel path 27c within the time required for passing to occur, which is calculated based on the relative positional relationship and relative speed between the vehicle 10 and the parked vehicle 12.

[0048] In S207, the vehicle speed of the host vehicle 10 during passing is calculated using, for example, Fig. 12 based on the host vehicle deceleration ratio determined according to the boundary height and the degree of contact influence using Fig. 7 and the difference between the road width between the left and right boundaries and the vehicle width of the host vehicle 10. In Fig. 12, the vehicle speed is calculated so that the greater the deceleration ratio, the slower the vehicle speed will be, and the greater the width of the left and right boundaries, the faster the vehicle speed will be. Note that if the road width between the left and right boundaries is smaller than the vehicle width of the host vehicle 10, this is the case where S205 returns No, and S207 is not executed.

[0049] By changing the vehicle speed according to the width of the left and right boundaries as well as the rate of deceleration, the vehicle will slow down as the obstacle moves and the road narrows, and will stop if there is no longer enough room to maneuver. Also, when passing through a narrow road or near an object that could damage the vehicle if it comes into contact with it, the vehicle will decelerate in accordance with the driver's intention.

[0050] In S208, the control gains for steering and braking are calculated according to the approach risk. When approaching an object with a high approach risk, the gains are increased to prevent overshooting the target path as much as possible. When approaching an object with a low approach risk, the gains are decreased to suppress sudden steering and braking.

[0051] If it is determined in S205 that the road width is not wide enough for the two vehicles to pass each other, S210 to S213 are executed in order. The processing contents of S210 and S211 will be explained using Figure 13. Figure 13 shows a scene in which the host vehicle 10 and the oncoming vehicle 11 are passing each other, and since the oncoming vehicle 11 is traveling near the center of the road 20, the host vehicle 10 has no choice but to stop and wait for the oncoming vehicle 11 to move out of the way.

[0052] If the sum of the vehicle width w10b of the host vehicle 10 excluding the door mirrors and the vehicle width w11 of the oncoming vehicle 11 is greater than the width w20 of the road, the host vehicle 10 cannot pass each other on this road even if the host vehicle 10 folds its door mirrors, so S210 determines No. Also, if the obstacle is a parked vehicle or fallen object, it cannot be expected that the obstacle will move and widen the road, so S210 determines No.

[0053] If the answer in S210 is Yes, the process determines in S211 whether the door mirrors need to be folded in order to pass each other. If the sum of the vehicle width w10a including the door mirrors of the host vehicle 10 and the vehicle width w11 of the oncoming vehicle 11 is smaller than the width w20 of the road, the host vehicle 10 can pass each other without folding the door mirrors, and the result in S211 is No.

[0054] Furthermore, if the sum of the vehicle width w10a of the host vehicle 10 including the door mirrors and the vehicle width w11 of the oncoming vehicle 11 is greater than the width w20 of the road, but the sum of the vehicle width w10b of the host vehicle 10 excluding the door mirrors and the vehicle width w11 of the oncoming vehicle 11 is smaller than the width w20 of the road, the oncoming vehicle 11 can pass by the host vehicle 10 if the host vehicle 10 folds its door mirrors, so S211 determines Yes.

[0055] If the determination in S211 is Yes, the door mirrors are folded in S212. Note that, since folding the door mirrors prevents the driver from visually checking, it is preferable to fold the door mirrors after stopping the vehicle. Instead of the door mirrors, a camera that can check the sides of the vehicle and a screen for displaying the image from this camera may be used, and the driver may check the sides using the image from this camera while the door mirrors are folded.

[0056] In S213, a position as close to the boundary on the left side as possible before the point where a collision with an obstacle is expected is calculated as the stopping position.

[0057] In S209, the content of the notification to the driver is determined based on the risk of approaching a road boundary and the distance to the road boundary, for example, using FIG. 14. The greater the risk of approaching a road boundary, the earlier and louder the warning is issued, urging the driver to be more careful. Also, the closer the distance to the road boundary, the louder the warning is issued, notifying the driver that contact with the road boundary is imminent. In this way, by changing the content of the notification based on the risk of approaching a road boundary and the distance to the road boundary, the driver can be notified of the location of objects that require particular attention.

[0058] As described above, the vehicle control device according to this embodiment can provide passing assistance by detecting and identifying obstacles and road boundaries in the recognition system calculation unit 105 based on information acquired by the external environment recognition unit 101, and then calculating a driving route and determining whether to issue a notification in the control system calculation unit 108 according to the risk of approaching the boundary.

[0059] According to the vehicle control device of this embodiment, when the host vehicle passes an oncoming vehicle or a parked vehicle, if the risk of approaching the road boundary is low, that is, if the impact of contact with the host vehicle, such as a plant or a tree, is small, the host vehicle is allowed to approach the road boundary by reducing the margin between the host vehicle and the road boundary.If the risk of approaching the road boundary is high, that is, if contact with the host vehicle, such as a guardrail or a utility pole, is large, that is, if the impact of contact with the host vehicle, such as a guardrail or a utility pole, is large, the margin between the host vehicle and the road boundary is increased, and a route is calculated that moves away from the host vehicle.

[0060] Furthermore, by determining the risk of approaching the vehicle and the road boundary in stages according to the height of the road boundary, the degree of approach can be adjusted according to the part of the vehicle that comes into contact with the road boundary. This makes it possible to realize an operation such as allowing the wheel part to come into contact with the boundary, but not the body part or mirror part.

[0061] Furthermore, the vehicle control device according to this embodiment adjusts the timing and magnitude of the warning issued to the driver depending on the risk of approaching a road boundary and the distance to the road boundary. This makes it possible to notify the driver of the presence of an object that poses a high risk of contact while passing another vehicle on a narrow road, thereby more appropriately alerting the driver.

[0062] As described above, the vehicle control device according to this embodiment can provide driving assistance that enables vehicles to pass each other on narrow roads.

[0063] The main features of this embodiment can be summarized as follows.

[0064] The processor of the vehicle control device acquires the degree of impact (contact impact) on the host vehicle due to contact between the host vehicle and a boundary structure corresponding to the type of boundary structure that constitutes the boundary of the area in which the host vehicle is traveling (FIG. 8). The processor acquires an approach risk indicating the risk when the host vehicle approaches the boundary structure corresponding to a combination of the height and impact of the boundary structure (FIG. 7). Based on the approach risk, the processor changes the distance (margin) at which the host vehicle can approach the boundary structure (FIG. 10) or changes the content of the notification to the driver of the host vehicle (FIG. 14).

[0065] This allows the vehicle to pass another vehicle while allowing contact with the hedge 24 (Figure 9). The driver can also know the degree of risk of contact in advance. As a result, the passing assistance on narrow roads can be improved.

[0066] In detail, the processor changes the content of the notification to the driver of the vehicle based on the approach risk and the distance from the vehicle to the boundary (FIG. 14).

[0067] This allows the driver to easily understand the extent of the risk of contact and the distance involved.

[0068] The influence degree (contact influence degree) indicates the influence on the host vehicle caused by contact between the boundary constituent and at least the mirror, body, and wheel of the host vehicle (FIGS. 7 and 8).

[0069] This allows for passing of vehicles in a manner that allows contact only between the mirror, body, or wheel.

[0070] As shown in FIG. 4, the processor determines the target trajectory by combining multiple approach risks obtained for multiple boundary structures (for example, a wall 21, a low step 23, and an oncoming vehicle 11).

[0071] This makes it possible to determine a target trajectory that suppresses unnecessary climbing over lane markings or low steps, for example, if the road is wide enough, and climbs over them only when necessary (Figures 2 and 4).

[0072] The processor acquires the deceleration rate of the host vehicle corresponding to the approach risk (FIG. 7). Based on the deceleration rate, the processor determines the target vehicle control amount to decelerate or stop the host vehicle (FIG. 12). This ensures safety when passing other vehicles.

[0073] Specifically, the processor determines the target vehicle control amount to decelerate or stop the host vehicle based on a combination of the difference between the road width and the vehicle width and the degree of deceleration (FIG. 12).

[0074] This allows the vehicle to be controlled in accordance with the driver's intention to slow down or stop when passing through narrow roads or near objects that could damage the vehicle body if hit (boundary structures with a high risk of approach).

[0075] If the processor determines that the host vehicle can pass the oncoming vehicle if it folds the door mirrors based on the approach risk, it determines the target vehicle control amount to fold the door mirrors and stop the host vehicle (S213, FIG. 6). This makes it as easy as possible for the oncoming vehicle to pass.

[0076] When the host vehicle approaches a boundary structure, the processor determines the target vehicle control amount so that the control gains for steering and braking of the host vehicle increase as the approach risk increases (S208, FIG. 6). This makes it possible to prevent the host vehicle's trajectory from overshooting the target route.

[0077] It should be noted that the present invention is not limited to the above-described embodiment and includes various modifications. For example, the above-described embodiment has been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to an embodiment having all of the described configurations.

[0078] Furthermore, some or all of the above-described configurations, functions, etc. may be implemented in hardware, for example, by designing them as integrated circuits. Furthermore, the above-described configurations, functions, etc. may be implemented in software by a processor interpreting and executing a program that implements each function. Information such as the programs, tables, and files that implement each function can be stored in a memory, a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD.

[0079] The embodiment of the present invention may have the following aspects.

[0080] (1) A vehicle control device including: an external environment information acquisition unit that acquires external environment information around the host vehicle; a boundary information acquisition unit that acquires, based on the external environment information, information on boundary structures that form the boundary of an area in which the host vehicle is traveling; an approach risk acquisition unit that acquires, based on the height and type of the boundary structures, an approach risk that indicates the risk including the degree of impact on the host vehicle due to contact between the boundary structure and the host vehicle when the host vehicle approaches the boundary structure; and a driving planning unit that changes the approachable distance of the host vehicle from the boundary structure based on the approach risk and the motion state of the host vehicle to plan a target trajectory and a target vehicle control amount of the host vehicle, or changes the content of notification to the driver of the host vehicle.

[0081] (2) The vehicle control device according to (1), wherein the degree of influence indicates an influence on the host vehicle caused by contact between the boundary configuration and at least a mirror, a body, and a wheel of the host vehicle.

[0082] (3) The vehicle control device according to (1), wherein the operation planning unit calculates the target trajectory by combining a plurality of the approach risks acquired for a plurality of the boundary configurations.

[0083] (4) The vehicle control device according to (1), wherein the driving planner calculates the target vehicle control amount so as to decelerate or stop the host vehicle when the approach risk is higher than a predetermined value.

[0084] (5) The vehicle control device according to (1), wherein, when it is determined based on the approach risk that the host vehicle can pass an oncoming vehicle if the host vehicle folds door mirrors, the driving planning unit calculates the target vehicle control amount to fold the door mirrors of the host vehicle and stop the host vehicle.

[0085] (6) The vehicle control device according to (1), wherein the driving planner calculates the target vehicle control amount such that, as the host vehicle approaches the boundary structure, the higher the approach risk, the larger the control gain for steering and braking of the host vehicle.

[0086] According to (1)-(6), an approach risk indicating the risk, including the degree of impact on the vehicle due to contact, is acquired according to the height and type of road boundary. This allows for a passing assistance function that, depending on the approach risk, allows for contact with road boundaries with little impact, such as hedges, and allows for driving over low bumps. Furthermore, by evaluating the approach risk according to the height of the road boundary, for example, when approaching the same type of boundary structure, a passing operation is realized that allows for contact with a tire or wheel while approaching it more closely than when approaching a painted vehicle body or a door mirror that is easily damaged. This allows for driving that is closer to the driver's own driving experience, by moving away from objects with a high impact taking into account the part of the vehicle that comes into contact with the boundary structure and allowing for contact with objects with a low impact.

[0087] DESCRIPTION OF SYMBOLS 10...Own vehicle 11...Oncoming vehicle 12...Parked vehicle 20...Runway 21, 22...Wall 23...Step 24...Hedge 25L, 25R...Boundary 27a...Driving route 27b...Transition curve 27c...Driving route 101...External environment recognition unit 102...Self-position recognition unit 103...High-precision map 104...Vehicle sensor 105...Recognition system calculation unit 106...Obstacle detection unit 107...Road boundary identification unit 108...Control system calculation unit 109...Approach risk calculation unit 110...Driving route calculation unit 111...Operation amount calculation unit 112...Notification control unit 113...Actuator 114...Driver notification device

Claims

1. A vehicle control device having a processor that acquires a degree of impact on the host vehicle caused by contact between the host vehicle and a boundary structure that constitutes the boundary of an area in which the host vehicle is traveling, corresponding to the type of the boundary structure, acquires an approach risk indicating the risk when the host vehicle approaches the boundary structure, corresponding to a combination of the height of the boundary structure and the degree of impact, and changes the approachable distance of the host vehicle from the boundary structure or changes the content of a notification to the driver of the host vehicle based on the approach risk.

2. A vehicle control device as described in claim 1, characterized in that the degree of influence indicates the influence on the host vehicle caused by contact between the boundary structure and at least a mirror, a body, and a wheel of the host vehicle.

3. A vehicle control device according to claim 1, characterized in that the processor determines a target trajectory by combining a plurality of the approach risks obtained for a plurality of the boundary structures.

4. A vehicle control device as described in claim 1, characterized in that the processor acquires a deceleration rate of the host vehicle corresponding to the approach risk, and determines a target vehicle control amount so as to decelerate or stop the host vehicle based on the deceleration rate.

5. A vehicle control device as described in claim 1, characterized in that, when it is determined based on the approach risk that the vehicle can pass an oncoming vehicle by folding the door mirrors, the processor determines a target vehicle control amount to fold the door mirrors and stop the vehicle.

6. A vehicle control device as described in claim 1, characterized in that the processor determines target vehicle control amounts so that, when the vehicle approaches the boundary structure, the higher the risk of approach, the larger the control gains for steering and braking of the vehicle.

7. A vehicle control device as described in claim 1, characterized in that the processor changes the content of the notification to the driver of the vehicle based on the approach risk and the distance from the vehicle to the boundary.

8. A vehicle control device according to claim 4, characterized in that the processor determines a target vehicle control amount so as to decelerate or stop the host vehicle based on a combination of the difference between the road width and the vehicle width and the rate of deceleration.

Images