Vehicle control device and autonomous driving system

Abstract

In the vehicle control device (100) of the present application, the poor road traveling section (1008) determines whether or not there is a poor road in a certain range ahead of the vehicle based on the position of the vehicle estimated by the position estimation section (1001) and the position of the poor road acquired by the map information acquisition section (1011), and determines whether or not the poor road is detected based on the road surface information acquired by the outside information acquisition section (1002). In a case where it is determined that there is a poor road in the certain range ahead of the vehicle based on the position of the vehicle estimated by the position estimation section (1001) and the position of the poor road acquired by the map information acquisition section (1011), and it is determined that the poor road is not detected based on the road surface information acquired by the outside information acquisition section (1002), the poor road miss detection determination section (1006) determines whether or not there is a miss of the detection of the poor road based on the position of the poor road and the detectable range of the stereo camera (200).

Description

Technical Field

[0001] This invention relates to a vehicle control device and an automatic driving system. This application claims priority to Japanese Patent Application No. 2021-153263, filed on September 21, 2021, the contents of which are incorporated herein by reference. Background Technology

[0002] To improve vehicle ride comfort, the vehicle's suspension needs to adapt to various road conditions. Recently, a technology has been developed that uses a stereo camera mounted on the windshield to detect the road conditions in front of the vehicle and adjusts the vehicle's suspension appropriately based on the detected road conditions, thereby improving ride comfort (in other words, handling stability and ride comfort).

[0003] For example, Patent Document 1 discloses a system comprising: a stereo sensor unit consisting of a stereo camera that generates image data; a computing unit that extracts relevant image portions from the image data based on future vehicle path data and calculates the road unevenness variation on the future vehicle path based on the generated image data; and a suspension control unit that generates a signal to adapt the vehicle's suspension based on the calculated road unevenness variation. Existing technical documents Patent documents

[0004] Patent Document 1: Japanese Patent Application Publication No. 2017-100705 Summary of the Invention The problem the invention aims to solve

[0005] However, when using a stereo camera, the following problems may occur due to the limitations of the camera's detection range. Specifically, such as... Figure 6 As shown, for example, when a stereo camera is mounted on the windshield, its detection range S in the vertical direction is limited due to factors such as camera performance, mounting height h, and the field of view θ of the camera in the vertical direction. Because of this detection range S, the detectable area of ​​the stereo camera in the forward direction of the vehicle depends on the lower limit (Deadlm1) and the upper limit (Deadlm2) of the detectable area. Therefore, in the case of uneven or rough roads, if the uneven road exists within the detectable area of ​​the stereo camera, it will be detected by the stereo camera.

[0006] But in, for example, like Figure 7AWhen following a vehicle as shown, if the distance between your vehicle and the leading vehicle is less than the upper limit of the detectable range (Deadlm2), even if a bad road exists within the detectable range, the stereo camera will fail to detect it due to the obstruction of the leading vehicle. If you drive alongside the leading vehicle in this state, it will be like... Figure 7B As shown, if the vehicle passes the lower limit of the detectable range (Deadlm1) when the bad road is not detected (meaning the vehicle's position on the bad road is less than Deadlm1), a problem of missing the bad road will occur. As a result, the vehicle will be driven on the bad road without any suspension adjustments.

[0007] This invention was made to solve such a technical problem, and its purpose is to provide a vehicle control device and an automatic driving system that can prevent the failure to detect poor road conditions. Technical means to solve the problem

[0008] The vehicle control device of the present invention is characterized by comprising: a position estimation unit that estimates the position of the vehicle; an external information acquisition unit that acquires external information including road surface information based on an image captured by an onboard camera positioned in front of the vehicle; a map information acquisition unit that acquires map information including the location of a bad road; a bad road driving unit that adjusts at least one of the vehicle's suspension and vehicle speed to drive on a bad road; and a bad road determination unit that determines whether there is a bad road within a certain range in front of the vehicle based on the vehicle position estimated by the position estimation unit and the location of the bad road acquired by the map information acquisition unit. The system detects both bad roads and road surface information obtained by the external information acquisition unit. A bad road detection omission determination unit determines whether a bad road has been detected based on the vehicle's position inferred by the position inference unit and the location of the bad road obtained by the map information acquisition unit, and determines that no bad road has been detected based on the road surface information obtained by the external information acquisition unit. In such cases, the bad road detection omission determination unit determines whether there is a missed bad road detection based on the location of the bad road obtained by the map information acquisition unit and the detectable range of the vehicle-mounted camera.

[0009] In the vehicle control device of the present invention, when it is determined that there is a bad road within a certain range ahead of the vehicle based on the vehicle's position inferred by the position inference unit and the location of the bad road obtained by the map information acquisition unit, and it is determined that no bad road has been detected based on the road surface information obtained by the external information acquisition unit, the bad road detection omission determination unit determines whether there is a missed bad road detection based on the location of the bad road obtained by the map information acquisition unit and the detectable range of the vehicle-mounted camera. In this way, even if the detectable range of the vehicle-mounted camera is blocked by a preceding vehicle, the missed detection of bad roads can be reliably prevented. The effects of the invention

[0010] According to the present invention, it is possible to prevent the missed detection of poor paths. Attached Figure Description

[0011] Figure 1 This is a block diagram illustrating a vehicle control device according to an embodiment. Figure 2 A flowchart illustrating the control processing of a vehicle control device. Figure 3 This is a flowchart illustrating the process for determining whether or not a faulty path has been missed. Figure 4 This diagram illustrates the speed adjustment for vehicles traveling on rough roads. Figure 5 This is a block diagram illustrating an implementation of an autonomous driving system. Figure 6 This is a diagram used to illustrate the detection range and detectable interval of a stereo camera. Figure 7A This is a diagram used to illustrate the existing problem. Figure 7B This is a diagram used to illustrate the existing problem. Detailed Implementation

[0012] Hereinafter, embodiments of the vehicle control device and autonomous driving system of the present invention will be described with reference to the accompanying drawings. In the following description, the directions and positions of up / down, left / right, and front / back are based on a vehicle equipped with the vehicle control device. Furthermore, in the following description, unless otherwise specifically mentioned, it is assumed that the vehicle cannot avoid poor roads (in other words, it cannot drive by bypassing poor roads).

[0013] [Regarding vehicle control devices] Figure 1 This is a block diagram illustrating a vehicle control device according to an embodiment. The vehicle control device 100 of this embodiment is mounted in a vehicle 10. Figure 1 The example shown is a passenger car, but vehicle 10 is not limited to passenger cars; it can also be a bus, truck, trailer, etc. In addition to the vehicle control device 100, vehicle 10 also includes a stereo camera 200, a gyroscope sensor 300, a suspension 400, etc.

[0014] The stereo camera 200 is equivalent to the "vehicle-mounted camera" described in the claims, and it captures images of the area in front of the vehicle 10. This stereo camera 200, for example, consists of a pair of left and right cameras arranged with their optical axes parallel to each other at a predetermined optical axis interval (baseline length), and is mounted on the windshield of the vehicle 10. Each pair of left and right cameras is composed of an image sensor such as a CMOS sensor and an optical lens. Furthermore, besides a stereo camera, the vehicle-mounted camera may also be a monocular camera or the like.

[0015] Gyroscope sensor 300 detects the acceleration and vibration of vehicle 10. Vehicle speed sensor 500 detects the speed of vehicle 10.

[0016] Suspension 400 is configured on each wheel of vehicle 10. Suspension 400 is, for example, one of an electromagnetic suspension, a semi-active suspension, a fully active suspension, and an air suspension. The electromagnetic suspension has an electric motor (linear motor, rotary motor) as an electric actuator; the semi-active suspension has a hydraulic damper (hydraulic damper, hydraulic shock absorber) as a hydraulic actuator capable of adjusting damping force; the fully active suspension has a drive cylinder (hydraulic cylinder, pneumatic cylinder) as a hydraulic or pneumatic actuator; and the air suspension has an air spring as a pneumatic actuator that also functions as a vehicle height adjustment device. This actuator (electric motor, hydraulic damper, drive cylinder, air spring, etc.) is controlled by vehicle control unit 100 (more specifically, the rough-road driving unit 1008 of vehicle control unit 100). In other words, the rough-road driving unit 1008 adjusts the suspension 400 by controlling the actuators (electric motor, hydraulic damper, drive cylinder, air spring, etc.).

[0017] The vehicle control device 100 is, for example, a microcomputer composed of a CPU (Central Processing Unit), ROM (Read Only Memory), and RAM (Random Access Memory). The CPU performs calculations, the ROM serves as a secondary storage device that records the program used for calculations, and the RAM serves as a temporary storage device that stores the calculation process and temporary control variables. The vehicle control device 100 performs various controls of the entire vehicle 10 by executing the stored program.

[0018] like Figure 1 As shown, the vehicle control device 100 of this embodiment includes a position inference unit 1001, an external information acquisition unit 1002, a gyroscope data acquisition unit 1003, a bad road determination unit 1004, a communication control unit 1005, a bad road missed detection determination unit 1006, a bad road detection distance maintenance unit 1007, a bad road driving unit 1008, a calibration unit 1009, a learning value storage unit 1010, a map information acquisition unit 1011, and a map database (map DB) 1012.

[0019] The position inference unit 1001 uses a satellite signal receiver (not shown) built into the vehicle 10 or the vehicle control device 100 to infer the position of the vehicle 10. Specifically, the position inference unit 1001 infers the position of the vehicle 10, i.e., latitude and longitude, by analyzing signals received by the satellite signal receiver from multiple satellites constituting the satellite navigation system. Alternatively, the position inference unit 1001 can also infer the position of the vehicle 10 based on the detection results of the gyroscope sensor 300.

[0020] The external information acquisition unit 1002 acquires external information based on images captured by the stereo camera 200. This external information includes, for example, road surface image data representing road conditions, image data of preceding and oncoming vehicles, and image data of traffic signs and lanes.

[0021] The gyroscope data acquisition unit 1003 is equivalent to the "vibration acquisition unit" described in the claims, and acquires the vibration of the vehicle 10 based on the detection data of the gyroscope sensor 300.

[0022] The poor road determination unit 1004 determines whether there is a poor road within a certain range (e.g., within 2 km) in front of the vehicle based on the position of the vehicle 10 inferred by the position inference unit 1001 and the position of the poor road obtained by the map information acquisition unit 1011. Furthermore, the poor road determination unit 1004 determines whether a poor road is detected in front of the vehicle 10 based on road surface image data obtained by the external information acquisition unit 1002. Additionally, if the poor road missed detection determination unit 1006 determines that a poor road has been missed, and the position of the poor road obtained by the external information acquisition unit 1002 is within a certain range in front of the vehicle 10, the poor road determination unit 1004 determines that there is a poor road on that driving surface. Moreover, a poor road, for example, refers to a road surface with unevenness, steps, protrusions, or other objects that cause vibrations when the vehicle travels on these objects to exceed a predetermined vibration threshold.

[0023] The communication control unit 1005 transmits and receives control information with an external server or the like via a communication device (not shown) installed in the vehicle 10. Specifically, the communication control unit 1005 controls the transmission and reception of vehicle speed and suspension adjustment learning values ​​stored in the learning value storage unit 1010 to an external server such as the cloud, and receives vehicle speed and suspension adjustment learning values ​​stored in the external server. Furthermore, upon receiving vehicle speed and suspension adjustment learning values ​​stored in the external server, the communication control unit 1005 stores these received learning values ​​in the learning value storage unit 1010.

[0024] If the bad road determination unit 1004 determines that there is a bad road within a certain range in front of the vehicle based on the position of the vehicle 10 inferred by the position inference unit 1001 and the position of the bad road obtained by the map information acquisition unit 1011, and if the bad road determination unit 1004 determines that no bad road has been detected based on the road surface information obtained by the external information acquisition unit 1002, then the bad road missed detection determination unit 1006 determines whether there is a missed bad road detection based on the position of the bad road obtained by the map information acquisition unit 1011 and the detectable range of the stereo camera 200. Furthermore, as mentioned above, the detectable range of the stereo camera 200 is the range between the lower limit (Deadlm1) and the upper limit (Deadlm2) of the detectable range (for example, refer to...). Figure 6 ).

[0025] The poor road detection distance maintenance unit 1007 maintains the distance to the preceding vehicle in such a way that the detection range of the stereo camera 200 is not blocked by the preceding vehicle.

[0026] The poor road driving unit 1008 adjusts at least one of the vehicle's suspension and vehicle speed to drive on poor roads. Specifically, the poor road driving unit 1008 adjusts the suspension 400 based on road surface image data acquired by the external information acquisition unit 1002 to drive on poor roads. Furthermore, if the poor road detection omission determination unit 1006 determines that there is a missed poor road detection, the poor road driving unit 1008 adjusts the suspension 400 of the vehicle 10 based on the suspension adjustment learning values ​​stored in the learning value storage unit 1010 to drive.

[0027] Furthermore, when the bad road determination unit 1004 determines that there is a bad road in a certain range ahead of the vehicle based on the position of the vehicle 10 inferred by the position inference unit 1001 and the position of the bad road obtained by the map information acquisition unit 1011, the bad road driving unit 1008 adjusts the vehicle speed and drives according to the learning value of the vehicle speed stored in the learning value storage unit 1010.

[0028] The calibration unit 1009 calibrates the vehicle speed based on the vehicle vibration acquired by the gyroscope data acquisition unit 1003 when driving on a poor road. Specifically, when the vehicle 10 is driving on a poor road, the calibration unit 1009 calculates the vibration of the vehicle 10 based on the gyroscope data acquired by the gyroscope data acquisition unit 1003. Then, if the calculated vibration of the vehicle 10 exceeds a preset threshold, it is determined that the vibration is large, and the vehicle speed for the next driving on the poor road is calibrated. Furthermore, the calibration unit 1009 associates the calibrated vehicle speed with the location of the poor road and stores it in the learning value storage unit 1010, thereby updating the learning value.

[0029] The learning value storage unit 1010 associates the vehicle speed calibrated by the calibration unit 1009 with the location of the poor road and stores it as a learning value for the vehicle speed. In addition, the learning value storage unit 1010 associates the suspension adjustment value adjusted by the poor road driving unit 1008 with the location of the poor road and stores it as a learning value for the suspension adjustment.

[0030] The map information acquisition unit 1011 acquires map information from the map database 1012. The map information includes the location of the inferior road, lane information, etc.

[0031] Below, according to Figure 2 The control processing of the vehicle control device 100 is explained. Figure 2 A flowchart illustrating the control processing of a vehicle control device.

[0032] First, in step S100, the map information acquisition unit 1011 obtains the location of the bad road from the map database 1012. At this time, the map information acquisition unit 1011 obtains the location of the bad road existing on the road surface of the driving path from the map database 1012, for example, based on the driving path of its own vehicle.

[0033] In step S101, following step S100, the malfunction determination unit 1004 determines whether there is a malfunction within a certain range (e.g., within 2 km) ahead based on the location of the malfunction obtained in step S100 and the position of the vehicle itself inferred by the position inference unit 1001. If no malfunction is determined, the control process returns to step S100. On the other hand, if a malfunction is determined, the control process proceeds to step S102.

[0034] In step S102, the poor road driving unit 1008 acquires the vehicle speed learning value stored in the learning value storage unit 1010. This vehicle speed learning value is obtained by associating the vehicle's previous driving on this driving path with the poor road position of that path. Furthermore, when the vehicle is driving on this driving path for the first time, the vehicle speed learning value is, for example, downloaded from an external server such as the cloud and stored in the learning value storage unit 1010 beforehand.

[0035] In step S103, following step S102, the vehicle control device 100 determines whether a speed adjustment is needed based on the current vehicle speed detected by the vehicle speed sensor 500 and the learned speed value obtained in step S102. Here, according to... Figure 4 The method for determining whether a speed adjustment is needed is explained.

[0036] Figure 4 In the example, 40 km / h represents a learning value for vehicle speed when driving on a poor road. Figure 4As shown, the vehicle control device 100 first calculates the distance to the bad road based on the location of the bad road obtained in step S100 and the location of its own vehicle inferred by the location inference unit 1001. Figure 4 For example, the distance to the poor road is 100m. Next, the vehicle control device 100 compares the current vehicle speed (e.g., 60km / h) with the learned speed value obtained in step S102 (here, the speed for driving on the poor road is 40km / h). If the comparison result shows that the current vehicle speed is greater than the learned speed, the vehicle control device 100 determines that the vehicle speed needs to be adjusted.

[0037] Then, the vehicle control unit 100 calculates the braking distance required to decelerate from the current vehicle speed to the learning value at a deceleration of less than a certain G. Figure 4 In this case, the braking distance is, for example, 50m. Then, the vehicle control device 100 determines the deceleration start position based on the calculated braking distance, working backwards from the position on the rough road. In this way, when driving on a rough road, the vehicle's suspension is adjusted to decelerate, thereby further suppressing vibrations when driving on rough roads compared to situations where only the suspension is adjusted, and avoiding sudden braking.

[0038] Subsequently, if it is determined in step S103 that no speed adjustment is needed, the control process proceeds to step S105. On the other hand, if it is determined that a speed adjustment is needed, the control process proceeds to step S104.

[0039] In step S104, the vehicle control device 100 adjusts the vehicle speed as described above by changing it to a learned value.

[0040] In step S105, following step S104, the poor road detection distance maintenance unit 1007 maintains a vehicle-to-vehicle distance such that the detectable range of the stereo camera 200 is not obscured by the preceding vehicle, and that this distance is at least equal to the upper limit of the detectable range (Deadlm2). Thus, by ensuring that the detectable range of the stereo camera 200 is not obscured by the preceding vehicle, missed detections of poor roads, etc., can be prevented.

[0041] In step S106, following step S105, the external information acquisition unit 1002 acquires road surface image data captured by the stereo camera 200.

[0042] In step S107, following step S106, the poor road determination unit 1004 determines whether a poor road has been detected based on the road surface image data obtained in step S106. Furthermore, known techniques can be used to detect poor roads based on road surface image data, so detailed explanations are omitted.

[0043] Subsequently, if a poor road condition is detected, the control process proceeds to step S108. In step S108, the poor road travel unit 1008 adjusts the suspension 400 based on the road surface image data obtained in step S106. Since known techniques can be used to adjust the vehicle's suspension based on the road surface image data, a detailed description is omitted. Then, when step S108 ends, the control process proceeds to step S111.

[0044] On the other hand, if it is determined in step S107 that no bad path has been detected, the control process proceeds to step S109.

[0045] In step S109, the faulty path omission determination unit 1006 determines whether there is a faulty path omission. The determination of whether there is a faulty path omission will be described later. Then, if a faulty path omission is determined, the control process proceeds to step S110. If no faulty path omission is determined, the control process proceeds to step S112.

[0046] In step S110, the rough road driving unit 1008 acquires the suspension adjustment learning values ​​stored in the learning value storage unit 1010, and adjusts the suspension 400 based on the acquired suspension adjustment learning values. These suspension adjustment learning values ​​are obtained by associating the vehicle's previous driving on this driving path with the rough road position of that driving path. Furthermore, when the vehicle is driving on this driving path for the first time, these suspension adjustment learning values ​​are, for example, downloaded from an external server such as the cloud and stored in the learning value storage unit 1010 beforehand.

[0047] In step S111, following step S108 or S110, the vehicle control device 100 determines whether the vehicle 10 has passed the bad road based on the bad road location obtained in step S100 and the current position of the vehicle itself inferred by the position inference unit 1001. If it is determined that the bad road has been passed, the control process proceeds to step S112.

[0048] In step S112, the poor road driving unit 1008 restores the vehicle speed and suspension 400, which were adjusted for driving on poor roads, to their original states.

[0049] In step S113, following step S112, the vehicle control device 100 determines whether the gyroscope data (i.e., vehicle vibration) acquired by the gyroscope data acquisition unit 1003 when driving on a poor road is greater than a preset threshold. If it is determined to be greater than the threshold, the control process proceeds to step S114. On the other hand, if it is determined to be less than the threshold, the control process proceeds to step S115.

[0050] In step S114, the calibration unit 1009 calibrates the vehicle speed when driving on the poor road, and associates the calibrated vehicle speed with the location of the poor road and stores it in the learning value storage unit 1010, thereby updating the learning value.

[0051] In step S115, the suspension adjustment value when driving on a rough road is associated with the location of the rough road and stored in the learning value storage unit 1010, thereby updating the learning value of the suspension adjustment. Thus, a series of control processes end.

[0052] Below, according to Figure 3 The determination and processing of step S109 above will be explained. Figure 3 This is a flowchart illustrating the process for determining whether or not a faulty path has been missed.

[0053] like Figure 3 As shown, in step S1091, the vehicle control device 100 obtains the distance (i.e., the inter-vehicle distance L1) with the preceding vehicle.

[0054] In step S1092, following step S1091, the vehicle control device 100 calculates the distance L2 from its own vehicle to the bad road based on the position of its own vehicle inferred by the position inference unit 1001 and the position of the bad road obtained in step S100.

[0055] In step S1093, following step S1092, the vehicle control device 100 acquires the detectable range of the stereo camera 200. At this time, the vehicle control device 100 acquires the lower limit (Deadlm1) and the upper limit (Deadlm2) of the detectable range, which are used to define the detectable range.

[0056] In step S1094, following step S1093, the vehicle control device 100 determines whether the poor road exists within the detectable range of the stereo camera 200 (i.e., whether the relationship Deadlm1 ≤ L2 ≤ Deadlm2 is satisfied). If it is determined that the poor road exists within the detectable range of the stereo camera 200, the control process proceeds to step S1095. On the other hand, if it is determined that the poor road does not exist within the detectable range of the stereo camera 200, the process of step S1094 is repeated.

[0057] In step S1095, following step S1094, the poor road determination unit 1004 determines whether a poor road has been detected based on the road surface image data acquired in step S106. If a poor road is detected, the control process proceeds to step S108. Conversely, if no poor road is detected, the control process proceeds to step S1096. Furthermore, the processing in step S1095 is the same as in step S107, and therefore can be omitted.

[0058] In step S1096, the vehicle control device 100 determines whether the detectable range of the stereo camera 200 is blocked by the preceding vehicle (i.e., whether the relationship L1≤L2 is satisfied).

[0059] Subsequently, when L1 > L2, the vehicle control device 100 determines that the detectable section is not obstructed by the preceding vehicle. Therefore, the control process proceeds to step S1099. In step S1099, the vehicle control device 100 determines that there are no missed detections of faulty roads and that faulty roads have been repaired. When step S1099 ends, the control process proceeds to step S112 as described above.

[0060] On the other hand, in step S1096, when L1≤L2, the vehicle control device 100 determines that the detectable area is blocked by the preceding vehicle. Therefore, the control process proceeds to step S1097. In step S1097, the vehicle control device 100 determines whether the poor road is outside the detectable area of ​​the stereo camera 200 (i.e., whether the relationship L2<Deadlm1 is satisfied).

[0061] Subsequently, when L2≥Deadlm1, the vehicle control device 100 determines that the poor road is not outside the detectable range of the stereo camera 200. Therefore, the control process returns to step S1095.

[0062] On the other hand, when L2 < Deadlm1, the vehicle control device 100 determines that the bad road is outside the detectable range of the stereo camera 200, and the control process proceeds to step S1098. In step S1098, the bad road miss detection determination unit 1006 determines that a bad road has been missed. As a result, the control process proceeds to the aforementioned step S110.

[0063] In the vehicle control device 100 configured as described above, if the malfunction determination unit 1004 determines that there is a malfunction within a certain range ahead of the vehicle based on the position of the vehicle 10 inferred by the position inference unit 1001 and the position of the malfunction obtained by the map information acquisition unit 1011, and determines that no malfunction is detected based on the road surface image data obtained by the external information acquisition unit 1002, then the malfunction missed detection determination unit 1006 determines whether there is a missed detection of a malfunction based on the position of the malfunction obtained by the map information acquisition unit 1011 and the detectable range of the stereo camera 200. In this way, even if the detectable range of the stereo camera 200 is blocked by a preceding vehicle, the missed detection of malfunctions can be reliably prevented.

[0064] Furthermore, this embodiment assumes that vehicle 10 cannot avoid the poor road, but vehicle 10 can also drive in a way that allows it to avoid the poor road. In this case, for example, map information acquisition unit 1011 acquires lane information contained in the map information, and vehicle control device 100 drives to a lane without poor roads based on the lane information acquired by map information acquisition unit 1011 and the location of the poor road. In this way, driving in a lane without poor roads can be achieved by changing to a lane without poor roads.

[0065] Furthermore, a path generation unit can be provided in the vehicle control device 100 to generate a driving route to the destination. This path generation unit generates a driving route that avoids the bad road based on the location of the bad road obtained by the map information acquisition unit 1011. In this way, a route that avoids the bad road can be generated, thus enabling driving that avoids the bad road.

[0066] [About autonomous driving systems] Figure 5 This is a block diagram illustrating an autonomous driving system according to an embodiment. The autonomous driving system 1 of this embodiment includes the vehicle 10 described above and a cloud 20 that is communicatively connected to the vehicle 10.

[0067] Cloud 20 is equivalent to the "external server" described in the claims, and has a communication unit 201 that communicates with the vehicle 10 and an information storage unit 202 that stores the learning values ​​of vehicle speed and suspension adjustment received by the communication unit 201.

[0068] The communication unit 201 transmits and receives information with the vehicle 10 via a network (not shown). The information storage unit 202 filters and stores the received vehicle speed learning value, suspension adjustment learning value, and vehicle 10 technical specifications.

[0069] In the autonomous driving system 1 configured as described above, the cloud 20 stores and accumulates learning values ​​of vehicle speed and suspension adjustment sent from multiple vehicles 10. Furthermore, it sends the stored learning values ​​of vehicle speed and suspension adjustment to the requesting vehicle, for example, upon request, thereby enabling these learning values ​​to be shared among multiple vehicles. Moreover, since the learning values ​​of vehicle speed and suspension adjustment are associated with the location of the rough road, even when other vehicles are driving on a rough road for the first time, they can drive based on these learning values, thus achieving excellent driving comfort.

[0070] The embodiments of the present invention have been described in detail above, but the present invention is not limited to the above embodiments, and various design changes can be made within the scope of the spirit of the present invention as described in the claims. Symbol Explanation

[0071] 1…Autonomous driving system, 10…Vehicle, 20…Cloud, 100…Vehicle control device, 200…Stereo camera (vehicle-mounted camera), 300…Gyroscope sensor, 400…Suspension, 500…Vehicle speed sensor, 1001…Position inference unit, 1002…External information acquisition unit, 1003…Gyroscope data acquisition unit (vibration acquisition unit), 1004…Poor road judgment unit, 1005…Communication control unit, 1006…Poor road missed detection judgment unit, 1007…Poor road detection distance maintenance unit, 1008…Poor road driving unit, 1009…Calibration unit, 1010…Learning value storage unit, 1011…Map information acquisition unit, 1012…Map DB.

Claims

1. A vehicle control device, characterized in that, have: The location inference unit infers the vehicle's position. The external information acquisition unit acquires external information, including road surface information, based on images captured by an onboard camera in front of the camera vehicle. The map information acquisition unit acquires map information including the location of the inferior road; The vehicle is used for driving on poor roads by adjusting at least one of the vehicle's suspension and speed. The bad road determination unit determines whether there is a bad road within a certain range in front of the vehicle based on the vehicle's position inferred by the position inference unit and the location of the bad road obtained by the map information acquisition unit, and determines whether a bad road is detected based on the road surface information obtained by the external information acquisition unit. as well as The poor road detection omission determination unit, when determining that there is a poor road within a certain range ahead of the vehicle based on the vehicle's position inferred by the position inference unit and the location of the poor road obtained by the map information acquisition unit, and determining that no poor road has been detected based on the road surface information obtained by the external information acquisition unit, determines whether there is a missed detection of poor road based on the location of the poor road obtained by the map information acquisition unit and the detectable range of the vehicle-mounted camera. When the lower limit of the detectable range of the vehicle-mounted camera in the direction ahead of the vehicle is set to Deadlm1, the upper limit of the detectable range is set to Deadlm2, the distance to the preceding vehicle is set to L1, and the distance from the vehicle itself to the bad road is set to L2, the bad road missed detection determination unit determines that there is a bad road missed detection when the relationship from Deadlm1≤L2≤Deadlm2 and L1≤L2 becomes L2<Deadlm1.

2. The vehicle control device according to claim 1, characterized in that, It includes a vibration acquisition unit and a calibration unit. The vibration acquisition unit acquires the vibration of the vehicle, and the calibration unit calibrates the vehicle speed based on the vehicle vibration acquired by the vibration acquisition unit when driving on a poor road.

3. The vehicle control device according to claim 1 or 2, characterized in that, The vehicle is equipped with a bad road detection distance maintenance unit, which maintains the distance to the preceding vehicle in such a way that the detectable range of the vehicle-mounted camera is not blocked by the preceding vehicle.

4. The vehicle control device according to claim 2, characterized in that, The system includes a learning value storage unit, which associates the vehicle speed calibrated by the calibration unit with the location of the poor road and stores it as a learning value for the vehicle speed.

5. The vehicle control device according to claim 4, characterized in that, When it is determined that there is a bad road within a certain range ahead of the vehicle based on the vehicle's position inferred by the position inference unit and the bad road position obtained by the map information acquisition unit, the bad road driving unit adjusts the vehicle speed according to the vehicle speed learning value stored in the learning value storage unit.

6. The vehicle control device according to claim 4, characterized in that, The learning value storage unit associates the suspension adjustment value after adjustment by the poor road driving unit with the position of the poor road and stores it as the suspension adjustment learning value.

7. The vehicle control device according to claim 6, characterized in that, If the poor road detection omission determination unit determines that there is a missed poor road detection, the poor road driving unit adjusts the vehicle's suspension according to the suspension adjustment learning value stored in the learning value storage unit.

8. The vehicle control device according to claim 6, characterized in that, The device includes a communication control unit that controls the transmission of vehicle speed learning values ​​and suspension adjustment learning values ​​stored in the learning value storage unit to an external server and the reception of vehicle speed learning values ​​and suspension adjustment learning values ​​stored in the external server.

9. The vehicle control device according to claim 1, characterized in that, The map information acquisition unit then acquires the lane information contained in the map information. The vehicle control device drives to a lane without bad roads based on the lane information obtained by the map information acquisition unit and the location of the bad road.

10. The vehicle control device according to claim 1, characterized in that, It includes a path generation unit that generates a driving path to the destination. The path generation unit generates a driving path that avoids the bad road based on the location of the bad road obtained by the map information acquisition unit.

11. An autonomous driving system comprising a vehicle and an external server, the vehicle being equipped with a vehicle control device according to any one of claims 6 to 8, the external server being configured to communicate with the vehicle, characterized in that... The external server has a communication unit and an information storage unit. The communication unit communicates with the vehicle, and the information storage unit stores the learned values ​​of vehicle speed and suspension adjustment received by the communication unit.

Images