Travel route creation device, travel route creation method, and non-transitory computer readable medium

Abstract

Disclosed is a travel route creation device, a travel route creation method, and a non-transitory computer-readable medium. The travel route creation device creates a travel route for a vehicle to travel along, from each of first data generated by a first sensor mounted on the vehicle, outputs first object information indicating a position of an object around the vehicle at a time of generation of the first data, and first object prediction information indicating a predicted position of the object around the vehicle at a time that is a predetermined time later than the time of generation of the first data, from each of second data generated by a second sensor mounted on the vehicle, outputs second object information indicating a position of an object around the vehicle, until a time that is the predetermined time before a change time at which an automatic driving level is changed, uses the first object information from a time that is the predetermined time before the change time to the change time, and uses the second object information after the change time.

Description

Technical Field

[0001] This disclosure relates to a driving path creation apparatus, a driving path creation method, and a computer program for creating driving paths. Background Technology

[0002] The driving path creation device uses data that is appropriately selected from multiple data points corresponding to the vehicle's surrounding conditions, which are output by multiple sensors mounted on the vehicle, to create a driving path that the vehicle should take.

[0003] Patent Document 1 describes a driving control device that continuously generates a target track for autonomous driving, regardless of the switching between autonomous and manual driving. The driving control device described in Patent Document 1 uses the generated target track to perform autonomous driving when switching from manual to autonomous driving.

[0004] Existing technical documents

[0005] Patent documents

[0006] Patent Document 1: International Publication No. 2018 / 073883 Summary of the Invention

[0007] In order to smoothly switch from one level of autonomous driving to another using the driving control device described in Patent Document 1, it is necessary to continuously generate the driving control data used in each level of autonomous driving before and after the switch. Therefore, the required hardware resources are increased when performing a smooth switch from one level of autonomous driving to another using the driving control device described in Patent Document 1.

[0008] The purpose of this disclosure is to provide a driving path creation device that can create driving paths corresponding to the switching of autonomous driving levels without excessively increasing hardware resources.

[0009] The driving path generation apparatus disclosed herein includes: a first detection unit that detects the positions of objects surrounding the vehicle at the time the first data is generated, based on each piece of first data representing the vehicle's surrounding conditions generated in a time sequence by a first sensor mounted on the vehicle, and outputs first object information representing those positions; and detects predicted positions of objects surrounding the vehicle at a time predetermined after the time the first data is generated, and outputs first object prediction information representing those predicted positions; and a second detection unit that detects objects surrounding the vehicle based on each piece of second data representing the vehicle's surrounding conditions generated in a time sequence by a second sensor mounted on the vehicle, which is different from the first sensor. The system determines the location of the vehicle and outputs second object information representing that location; the prediction unit predicts whether the autonomous driving level controlling the vehicle's movement will change from a first autonomous driving level where the driver has a first monitoring duty to a second autonomous driving level where the driver has a second monitoring duty different from the first monitoring duty, and if the prediction indicates a change, determines the time of the predicted change; and the production unit uses the first object information to produce a driving path for the vehicle up to a time predetermined before the time of change, uses the first object prediction information to produce a driving path from the time predetermined before the time of change until the time of change, and uses the second object information to produce a driving path after the time of change.

[0010] In the driving path creation apparatus disclosed herein, it is preferable that the first detection unit calculates the reliability of the first object prediction information and outputs confidence information representing the reliability, and the creation unit uses the first object prediction information and confidence information to create a driving path from a time predetermined before the change time to the change time.

[0011] In the driving path creation apparatus disclosed herein, it is preferable that the creation unit uses first object prediction information corresponding to the confidence information to create a driving path when the reliability indicated in the confidence information exceeds the confidence threshold, and does not use the first object prediction information corresponding to the confidence information to create a driving path when the reliability indicated in the confidence information does not exceed the confidence threshold.

[0012] In the driving path making apparatus disclosed herein, it is preferable that the operation state of the first detection unit changes from open to closed, and the operation state of the second detection unit changes from closed to open, at a time predetermined before the change time.

[0013] In the driving path generation apparatus disclosed herein, it is preferable that the first detection unit uses the first data to output the first object information and the first object prediction information when the generation time of the first data is more than a predetermined time before the change time, and does not use the first data to output the first object information and the first object prediction information when the generation time of the first data is not more than a predetermined time before the change time. Similarly, the second detection unit does not use the second data to output the second object information when the generation time of the second data is more than a predetermined time before the change time, and uses the second data to output the second object information when the generation time of the second data is not more than a predetermined time before the change time.

[0014] In the driving path generation apparatus disclosed herein, it is preferable that the first detection unit uses a peripheral image generated in a time sequence by a first sensor based on light imaged by an optical system having a first focal length as first data, and outputs first object information and first object prediction information, and the second detection unit uses a peripheral image generated in a time sequence by a second sensor based on light imaged by an optical system having a second focal length shorter than the first focal length as second data, and outputs second object information.

[0015] The driving path generation method disclosed herein includes: detecting the positions of objects surrounding the vehicle at the time of generation of each piece of first data representing the vehicle's surrounding conditions, generated in a time series by a first sensor mounted on the vehicle, and outputting first object information representing those positions; and detecting the predicted positions of objects surrounding the vehicle at a predetermined time after the generation time of the first data, and outputting first object prediction information representing those positions; and detecting the predicted positions of objects surrounding the vehicle at a predetermined time after the generation time of the first data, and outputting first object prediction information representing those positions; and detecting the positions of objects surrounding the vehicle at the time series of each piece of second data representing the vehicle's surrounding conditions, generated in a time series by a second sensor mounted on the vehicle (different from the first sensor). The system sets the location and outputs second object information representing that location; it predicts whether the autonomous driving level controlling the vehicle's movement will change from the first autonomous driving level where the driver has a first monitoring duty to the second autonomous driving level where the driver has a second monitoring duty different from the first monitoring duty; if the prediction is a change, it determines the time of the predicted change; and until a time predetermined before the time of change, it uses the first object information to create a driving path for the vehicle; from the time predetermined before the time of change until the time of change, it uses the first object prediction information to create a driving path; and after the time of change, it uses the second object information to create a driving path.

[0016] The driving path creation computer program disclosed herein, stored on a non-transitory computer-readable medium, causes a computer mounted in a vehicle to execute: from each of the first data representing the vehicle's surrounding conditions generated in a time series by a first sensor mounted in the vehicle, detecting the positions of objects surrounding the vehicle at the time the first data was generated, and outputting first object information representing those positions; and detecting the predicted positions of objects surrounding the vehicle at a predetermined time later than the time the first data was generated, and outputting first object prediction information representing those positions; and from the second data representing the vehicle's surrounding conditions generated in a time series by a second sensor mounted in the vehicle (different from the first sensor),... Each of the components detects the position of objects around the vehicle and outputs second object information representing that position; predicts whether the autonomous driving level controlling the vehicle's driving will change from the first autonomous driving level where the driver has a first monitoring duty to the second autonomous driving level where the driver has a second monitoring duty different from the first monitoring duty; if the prediction is a change, it determines the time of the predicted change; and until a time predetermined before the time of change, it uses the first object information to create a driving path for the vehicle; from the time predetermined before the time of change until the time of change, it uses the first object prediction information to create a driving path; and after the time of change, it uses the second object information to create a driving path.

[0017] The driving path creation device disclosed herein can create driving paths corresponding to the switching of autonomous driving levels without excessively increasing hardware resources. Attached Figure Description

[0018] Figure 1 This is a schematic structural diagram of a vehicle equipped with a driving path creation device.

[0019] Figure 2 This is a hardware structure diagram of the ECU.

[0020] Figure 3 This is a functional block diagram of the processor in an ECU.

[0021] Figure 4 This is a diagram illustrating an example of creating a driving path using the first and second data.

[0022] Figure 5 This is a flowchart of the process for creating and processing the driving route.

[0023] (Symbol Explanation)

[0024] 1: Vehicle; 5: ECU; 531: First Inspection Department; 532: Second Inspection Department; 533: Prediction Department; 534: Manufacturing Department; 535: Driving Control Department. Detailed Implementation

[0025] Hereinafter, with reference to the accompanying drawings, a driving path creation device capable of creating driving paths corresponding to changes in autonomous driving levels without excessively increasing hardware resources will be described in detail. The driving path creation device predicts whether the autonomous driving level controlling the vehicle's movement will change from a first autonomous driving level where the driver has a first monitoring duty to a second autonomous driving level where the driver has a second monitoring duty different from the first monitoring duty. If a change is predicted, the device determines the predicted change time. Furthermore, the driving path creation device uses first object information to create the vehicle's driving path up to a time predetermined before the change time; uses first object prediction information to create the driving path from the time predetermined before the change time up to the change time; and uses second object information to create the driving path after the change time. The first object information is information indicating the positions of objects surrounding the vehicle at the time the first data was generated, detected from first data representing the vehicle's surrounding conditions generated in a time sequence by a first sensor mounted on the vehicle. The first object prediction information is information indicating the predicted positions of objects surrounding the vehicle at a time predetermined after the generation time of the first data, detected from the first data. In addition, the second object information is information indicating the position of objects around the vehicle, which is generated in a time series from the second data representing the surrounding conditions of the vehicle by a second sensor that is different from the first sensor mounted on the vehicle.

[0026] Figure 1 This is a schematic structural diagram of a vehicle equipped with a driving path creation device.

[0027] Vehicle 1 includes a peripheral camera 2, a GNSS receiver 3, a storage device 4, and an ECU 5 (Electronic Control Unit). ECU 5 is an example of a driving path creation device. The peripheral camera 2, GNSS receiver 3, storage device 4, and ECU 5 are communicatively connected via an in-vehicle network according to a standard controller area network.

[0028] The peripheral camera 2 has a two-dimensional detector, such as a CCD or C-MOS, composed of an array of photoelectric conversion elements sensitive to visible light, and an imaging optical system that images the area to be photographed onto the two-dimensional detector. The peripheral camera 2 is positioned, for example, at the upper front of the vehicle interior, facing forward, and photographs the surrounding conditions of the vehicle 1 through the windshield at predetermined shooting cycles (e.g., 1 / 30 to 1 / 10 of a second), outputting a peripheral image representing the surrounding conditions as first data or second data. The peripheral camera 2 includes a telescopic peripheral camera 2-1 and a wide-angle peripheral camera 2-2.

[0029] The telescopic peripheral camera 2-1 and the wide-angle peripheral camera 2-2 are examples of a first sensor and a second sensor, respectively, used to detect the surrounding conditions of the vehicle 1. The imaging optical system of the telescopic peripheral camera 2-1 has a longer focal length than the imaging optical system of the wide-angle peripheral camera 2-2. The peripheral camera 2 may also have three or more cameras.

[0030] As either the first or second sensor, the vehicle may also have a LiDAR (Light Detection and Ranging) sensor or a RADAR (Radio Detection and Ranging) sensor. Based on the surrounding conditions of the vehicle 1, the LiDAR or RADAR sensor outputs a distance image, where each pixel has a value corresponding to the distance to the object represented by that pixel, as the first or second data.

[0031] GNSS receiver 3 receives GNSS signals from GNSS (Global Navigation Satellite System) satellites at predetermined intervals, and determines the vehicle 1's own position based on the received GNSS signals. GNSS receiver 3 then outputs a positioning signal, representing the vehicle 1's own position based on the GNSS signals, to ECU 5 via the in-vehicle network at predetermined intervals.

[0032] Storage device 4 is an example of a storage unit, such as a hard disk drive or a non-volatile semiconductor memory. Storage device 4 stores map data, including information related to road features such as lane markings, in a location-corresponding manner.

[0033] ECU5 uses map data of the surrounding area corresponding to the position measured by the GNSS receiver 3 to determine the timing of changing the autonomous driving level. Based on the time interval up to the time of change, it selects data output by the surrounding cameras 2 to create a driving path for vehicle 1. Additionally, ECU5 performs driving control processing by sending control signals to driving mechanisms such as the engine, motor, brakes, and steering wheel to cause vehicle 1 to travel along the created driving path. Alternatively, vehicle 1 may also have a different ECU than ECU5, through which driving control processing is performed.

[0034] Figure 2 This is a hardware schematic diagram of ECU5. ECU5 includes a communication interface 51, a memory 52, and a processor 53.

[0035] Communication interface 51 is an example of a communication unit, having a communication interface circuit for connecting ECU 5 to an in-vehicle network. Communication interface 51 supplies received data to processor 53. Additionally, communication interface 51 outputs data supplied from processor 53 to an external source.

[0036] The memory 52 includes both volatile and non-volatile semiconductor memory. The memory 52 stores various data used in the processing of the processor 53, such as information indicating a predetermined time for changing information used in the creation of the driving path based on time intervals up to the moment when the autonomous driving level is changed. Additionally, the memory 52 temporarily stores the positions of surrounding objects at the time the surrounding images detected from the surrounding images are generated. Furthermore, the memory 52 stores various application programs, such as a driving path creation program that performs the driving path creation processing.

[0037] Processor 53 is an example of a control unit, having one or more processors and their peripheral circuitry. Processor 53 may also have other arithmetic circuitry such as a logic unit, a numerical arithmetic unit, or a graphics processing unit.

[0038] Figure 3 This is a functional block diagram of the processor 53 in ECU5. Figure 4 This is a diagram illustrating an example of creating a driving path using the first and second data.

[0039] In the processor 53 of the ECU 5, there are functional blocks including a first detection unit 531, a second detection unit 532, a prediction unit 533, a production unit 534, and a driving control unit 535. These units of the processor 53 are functional modules installed via a computer program stored in the memory 52 and executed on the processor 53. The computer program that implements the functions of each unit of the processor 53 may also be provided in the form of a computer-readable, portable recording medium such as a semiconductor memory, magnetic recording medium, or optical recording medium. Alternatively, these units of the processor 53 may also be installed in the ECU 5 as independent integrated circuits, microprocessors, or firmware.

[0040] The first detection unit 531 detects the position of the surrounding objects of the vehicle 1 at the time of generation of each of the surrounding images generated by the telescopic peripheral camera 2-1 in a time sequence, and outputs the first object information representing the position.

[0041] The first detection unit 531 detects other vehicles in the vicinity of vehicle 1 by inputting the peripheral image output by the telescopic peripheral camera 2-1 into a recognizer that has been pre-learned in order to detect predetermined objects represented in the peripheral image.

[0042] The recognizer can be configured, for example, as a convolutional neural network (CNN) with multiple convolutional layers connected in series from the input side to the output side. The CNN is pre-learned by using multiple images representing predetermined objects to be detected as teacher data inputs to the CNN, thereby acting as a recognizer for detecting objects.

[0043] The first detection unit 531 uses the current position and posture of the vehicle 1, the direction from the vehicle 1 toward the surrounding objects, and the estimated distance from the vehicle 1 to the surrounding objects to estimate the actual spatial position of the detected surrounding objects, and outputs the first object information.

[0044] The first detection unit 531 determines the current position of the vehicle 1 based on the positioning signal obtained from the GNSS receiver 3. Additionally, the first detection unit 531 can also determine the vehicle 1's posture by detecting ground features such as lane markings from surrounding images and matching the detected ground features with corresponding ground features represented in the map data stored in the storage device 4.

[0045] The first detection unit 531 uses the position of the surrounding object on the peripheral image, the focal length of the imaging optical system of the telescope peripheral camera 2-1, and the tilt of the optical axis of the imaging optical system of the telescope peripheral camera 2-1 relative to the direction of travel of the vehicle 1 to infer the direction from the vehicle 1 toward the surrounding object.

[0046] The first detection unit 531 estimates the distance from the vehicle 1 to the surrounding objects based on internal parameters such as the reference size of the surrounding objects stored in the memory 52, the size of the surrounding objects represented in the surrounding image, the focal length of the imaging optical system of the telescopic peripheral camera 2-1, and the pixel size of the surrounding image.

[0047] The first detection unit 531 detects the predicted positions of objects surrounding the vehicle 1 at a predetermined time after the generation time of each of the surrounding images generated by the telescopic peripheral camera 2-1 in a time sequence, and outputs the first object prediction information representing the predicted position.

[0048] The first detection unit 531 tracks surrounding objects detected from a series of surrounding images output in a time sequence by the telescopic peripheral camera 2-1, and detects the predicted positions of the surrounding objects from the trajectories obtained through the tracking results. The first detection unit 531 performs tracking processing using a Kalman filter or particle filter, etc., on the positions of the surrounding objects detected from the series of surrounding images at the time of image generation, thereby enabling the tracking of surrounding objects. In the tracking processing, odometer information representing the current position of vehicle 1, inferred from the rotation angle of the wheels of vehicle 1, can also be used.

[0049] The first detection unit 531 can also calculate the reliability of the predicted position represented in the first object prediction information and output confidence information representing that reliability. For example, the first detection unit 531 outputs confidence information in a manner that the smaller the variance of the prediction error of the predicted position in the tracking process, the greater the reliability of the predicted position.

[0050] The first detection unit 531 can also detect the predicted position of the surrounding objects by inputting the positions of the surrounding objects at the time of generation of the surrounding images detected from a series of surrounding images into a recognizer using a neural network. The recognizer pre-learns the positions of the surrounding objects, the actual spatial positions of the objects after a predetermined time, the position of the camera when each image was generated, the optical axis direction, and the distance from the camera to the predetermined object as teacher data. In addition, the recognizer can also output the reliability of the predicted position along with the predicted position of the surrounding objects.

[0051] The first detection unit 531 can also detect the predicted position of the surrounding objects based on the amount of movement of the positions of the surrounding objects represented in the first object information output according to the time sequence.

[0052] exist Figure 4 In this system, the peripheral camera 2-1 generates peripheral images D1-1 to D1-5 in a time series, discretely timed intervals t1-t5. The first detection unit 531 outputs first object information O1-1 representing the position of an object from the peripheral image D1-1 generated at time t1. Additionally, the first detection unit 531 outputs first object prediction information OP1-1 representing the predicted position of the object at time t2, one unit later than the time t1 when the peripheral image D1-1 was generated. That is, in... Figure 4 In the example, the scheduled time is 1 unit of time.

[0053] exist Figure 4 The part with the network cable attached indicates that no data has been generated or no data processing has been performed (off state). Figure 4 In the example described later, prediction unit 533 determines time t4 as the change time. At time t3, a predetermined time before the change time t4, the first detection unit 531 changes its operation state from on to off, and subsequent information about the first object and prediction information about the first object are not output. The telescopic peripheral camera 2-1, which generates the peripheral image used in the detection processing of the first detection unit 531, can also change its operation state from on to off at time t3, together with the first detection unit 531.

[0054] The second detection unit 532 detects the position of the surrounding objects of the vehicle 1 at the time of generation of each of the surrounding images generated by the wide-angle peripheral camera 2-2 in a time sequence, and outputs the second object information representing the position.

[0055] The second detection unit 532 detects the positions of other vehicles and other objects in the vicinity of vehicle 1 by inputting the peripheral image output by the wide-angle peripheral camera 2-2 into a recognizer that has been pre-learned in order to detect predetermined objects represented in the peripheral image.

[0056] The recognizer can be set up as a CNN, for example. It is pre-learned by using multiple images representing a predetermined object to be detected as teacher data input into the CNN, which acts as a recognizer for detecting objects.

[0057] exist Figure 4 In the process, the wide-angle peripheral camera 2-2 generates peripheral images D2-1 to D2-5 in a time sequence from time t1 to t5. The second detection unit 532 outputs second object information O2-3, representing the position of the object, from the peripheral image D2-3 generated at time t3.

[0058] The second detection unit 532 changes its operating state from off to on at time t3, a predetermined time before the change time t4, and begins outputting the second object information. The wide-angle peripheral camera 2-2, which generates the peripheral image used in the detection processing of the second detection unit 532, can also have its operating state off before time t3, and change from off to on at time t3 together with the second detection unit 532.

[0059] The prediction unit 533 predicts whether the autonomous driving level of the vehicle 1 has been changed, and if the prediction is that it has been changed, it determines the time of the change.

[0060] The prediction unit 533 obtains the current level of autonomous driving from the driving control unit 535, which will be described later. The level of autonomous driving includes Level 1, where the driver has a duty to monitor the surroundings, and Level 2, where the driver does not have a duty to monitor the road ahead.

[0061] For example, if the vehicle is traveling at 80 km / h at Level 1 autonomous driving, where the driver has a duty to monitor the surroundings, and the speed decreases to 30 km / h due to congestion, the system can switch from Level 1 to Level 2 autonomous driving, where the driver does not have a duty to monitor the road ahead. In this case, the prediction unit 533 obtains traffic information indicating traffic conditions such as congestion and traffic control along the driving route via a data communication module (not shown), a wireless base station (not shown), and a communication network (not shown) connected to the in-vehicle network.

[0062] Based on the acquired traffic information, the prediction unit 533 estimates the location where control will end under the current level of autonomous driving. Furthermore, the prediction unit 533 adds the time obtained by dividing the distance from the current location to the location where control will end based on the current level of autonomous driving by the standard speed used in control based on that level of autonomous driving to the current time, thereby determining the change time. Figure 4 In the example, the prediction unit 533 determines the change time as t4.

[0063] Additionally, the prediction unit 533 can also obtain map data of the surrounding area of ​​the location corresponding to the positioning signal output by the GNSS receiver 3 from the storage device 4. The prediction unit 533 determines the end point of the drivable lane under the current level of autonomous driving, and adds the time obtained by dividing the distance from the current location to that location by the standard speed in the control based on the level of autonomous driving to the current time, thereby determining the change time.

[0064] Production department 534 uses information indicating the position of objects detected from the surrounding data generated by the surrounding camera 2 to create the driving path of vehicle 1.

[0065] The production unit 534 uses the positions of surrounding objects temporarily stored in the memory 52 at the previous moment and the first object information or first object prediction information output by the first detection unit 531 or the second object information output by the second detection unit 532 to infer the predicted position of future objects.

[0066] Production department 534 creates the driving path of vehicle 1 by ensuring a predetermined interval based on the predicted position of objects in the future.

[0067] Production unit 534 uses the first object information to create a driving path up to a predetermined time before the change time. Production unit 534 uses the first object prediction information to create a driving path from a time predetermined time before the change time until the change time. After the change time, production unit 534 uses the second object information to create a driving path.

[0068] exist Figure 4 In the example, the production unit 534 uses the first object information to create the driving path until one unit of time before the change time t4, that is, until it becomes time t3. For example, the production unit 534 uses the first object information O1-2 corresponding to the peripheral image D1-2 generated by the telescopic peripheral camera 2-1 to create the driving path DP-2 at time t2.

[0069] From time t3, which is one unit time earlier than the change time t4, to the change time t4, the production unit 534 uses the first object prediction information OP1-2 corresponding to the peripheral image D1-2 generated by the telescope peripheral camera 2-1 to produce the driving path DP-3.

[0070] The production unit 534 can also use the first object prediction information OP1-2 and the confidence information representing the reliability of the first object prediction information OP1-2 to create the driving path DP-3. For example, if the reliability represented by the confidence information exceeds a predetermined confidence threshold, the production unit 534 assumes that the surrounding object exists at the position indicated in the first object prediction information OP1-2 and creates the driving path DP-3. Conversely, if the reliability represented by the confidence information does not exceed the predetermined confidence threshold, the production unit 534 assumes that the surrounding object does not exist at the position indicated in the first object prediction information OP1-2 and does not use the first object prediction information OP1-2 to create the driving path DP-3.

[0071] After the change of time t4, the production unit 534 uses the second object information to create a driving path. For example, the production unit 534 uses the second object information O2-4 corresponding to the peripheral image D2-4 generated by the wide-angle peripheral camera 2-2 to create the driving path DP-4 at time t4.

[0072] Figure 5 This is a flowchart of the driving path creation process. ECU5 repeatedly executes the driving path creation process at predetermined time intervals while vehicle 1 is driving under automatic driving control.

[0073] First, the first detection unit 531 of ECU5 detects the position of the surrounding objects at the time of generation of each of the surrounding images generated in a time sequence by the telescopic peripheral camera 2-1, and outputs first object information indicating the position. In addition, the first detection unit 531 detects the predicted position of the surrounding objects at a time that is a predetermined time later than the time of generation of the surrounding images and outputs first object prediction information indicating the predicted position (step S1).

[0074] The prediction unit 533 of ECU5 predicts whether the autonomous driving level of vehicle 1 will change from the first autonomous driving level to the second autonomous driving level (step S2). If no change in the autonomous driving level is predicted (step S2: "No"), the processing of ECU5 proceeds to step S5 described later. If a change in the autonomous driving level is predicted (step S2: "Yes"), the prediction unit 533 determines the time of change of the autonomous driving level (step S3).

[0075] The production unit 534 of ECU5 determines whether the interval from the current time to the change time is within a predetermined time, that is, whether the current time is a time earlier than the change time by a predetermined time (step S4). If it is determined that the interval from the current time to the change time is not within a predetermined time (step S4: "No"), the production unit 534 uses the first object information to create a driving path (step S5), and the processing of ECU5 returns to step S1.

[0076] If it is determined that the interval from the current time to the change time is within the predetermined time (step S4: "Yes"), the production unit 534 uses the first object prediction information to create a driving path (step S6).

[0077] The second detection unit 532 of the ECU detects the position of the surrounding objects from each of the surrounding images generated in a time sequence by the wide-angle peripheral camera 2-2, and outputs the second object information representing the position (step S7).

[0078] After the production department 534 determines whether the current time has passed the change time, that is, whether the current time is a change time (step S8), if it is determined that the current time has not passed the change time (step S8: "No"), the processing of ECU5 returns to step S6.

[0079] If it is determined that the current time has changed (step S8: "Yes"), the production unit 534 uses the second object information to create a driving path (step S9) and ends the driving path creation process.

[0080] If a driving path creation process is executed again after the previous driving path creation process has ended, the second detection unit in the previously executed driving path creation process can function as the first detection unit in the later driving path creation process. Through this operation, ECU5 can appropriately create driving paths even when the autonomous driving level changes from Level 1 to Level 2 and then to other levels.

[0081] By performing this driving path creation process, ECU5 can create driving paths corresponding to the switching of autonomous driving levels without excessively increasing hardware resources.

[0082] It is hoped that those skilled in the art will understand that various changes, substitutions and modifications can be made without departing from the spirit and scope of this disclosure.

Claims

1. A travel path creation device, comprising: The first detection unit detects the position of the surrounding objects of the vehicle at the time of generation of the first data from one of the first data representing the surrounding conditions of the vehicle generated in a time sequence by the first sensor mounted on the vehicle, and outputs the first object information representing the position. It also detects the predicted position of the surrounding objects of the vehicle at a time that is a predetermined time later than the time of generation of the first data, and outputs the first object prediction information representing the predicted position. The second detection unit detects the position of objects around the vehicle from one of a plurality of second data representing the surrounding conditions of the vehicle, generated in a time sequence by a second sensor that is different from the first sensor mounted on the vehicle, and outputs second object information representing that position. The prediction unit predicts whether the autonomous driving level controlling the vehicle's operation will change from a first autonomous driving level where the driver has a first monitoring duty to a second autonomous driving level where the driver has a second monitoring duty different from the first monitoring duty. If the prediction indicates a change, the unit determines the time of the predicted change. as well as The production department uses the first object information to create a driving path for the vehicle up to a time earlier than the predetermined time of the change. From the time earlier than the predetermined time of the change until the time of the change, the production department uses the first object prediction information to create the driving path. After the time of the change, the production department uses the second object information to create the driving path.

2. The travel path creation device according to claim 1, wherein, The first detection unit calculates the reliability of the first object prediction information and outputs confidence information representing that reliability. The production unit uses the first object prediction information and the confidence information to create the driving path from the time before the predetermined time of change to the time of change.

3. The travel path creation device according to claim 2, wherein, If the reliability indicated in the confidence information exceeds the confidence threshold, the production unit uses the first object prediction information corresponding to the confidence information to produce the driving path; if the reliability indicated in the confidence information does not exceed the confidence threshold, the production unit does not use the first object prediction information corresponding to the confidence information to produce the driving path.

4. The travel path creation device according to any one of claims 1-3, wherein, At a time earlier than the predetermined time of change, the operation state of the first detection unit changes from on to off, and the operation state of the second detection unit changes from off to on.

5. The travel path creation device according to claim 4, wherein, The first detection unit uses the first data to output the first object information and the first object prediction information when the generation time of the first data is more than a predetermined time before the change time; however, it does not use the first data to output the first object information and the first object prediction information when the generation time of the first data is not more than a predetermined time before the change time. The second detection unit does not use the second data to output the information of the second object when the generation time of the second data is more than a predetermined time before the change time; and it uses the second data to output the information of the second object when the generation time of the second data is not more than a predetermined time before the change time.

6. The travel path creation apparatus according to any one of claims 1-3, wherein, The first detection unit outputs the first object information and the first object prediction information by using the surrounding image generated by the first sensor in a time series based on the light imaged by the optical system with the first focal length as the first data. The second detection unit outputs information about the second object by using the surrounding image generated by the second sensor in a time sequence based on light imaged by an optical system having a second focal length shorter than the first focal length as the second data.

7. A method for creating a driving path, comprising: From one of a plurality of first data representing the surrounding conditions of the vehicle generated in a time sequence by the first sensor mounted on the vehicle, the position of the surrounding objects of the vehicle at the time of generation of the first data is detected and the first object information representing the position is output. Furthermore, the predicted position of the surrounding objects of the vehicle at a time that is a predetermined time later than the time of generation of the first data is detected and the first object prediction information representing the predicted position is output. The position of an object in the vicinity of the vehicle is detected from one of a plurality of second data representing the surrounding conditions of the vehicle, generated in a time sequence by a second sensor different from the first sensor mounted on the vehicle, and second object information representing that position is output. Predict whether the autonomous driving level controlling the vehicle's driving will change from a first autonomous driving level where the driver has a first monitoring duty to a second autonomous driving level where the driver has a second monitoring duty different from the first monitoring duty. If the prediction is that the vehicle will change, determine the time of the predicted change. as well as The first object information is used to create the driving path of the vehicle until the time of change is earlier than the predetermined time. The first object prediction information is used to create the driving path from the time of change to the time of change. After the time of change, the second object information is used to create the driving path.

8. A non-transitory computer-readable medium storing a computer program for creating a driving route, the computer program for creating a driving route causing a computer mounted in a vehicle to execute: From one of a plurality of first data representing the surrounding conditions of the vehicle generated in a time sequence by the first sensor mounted on the vehicle, the position of the surrounding objects of the vehicle at the time of generation of the first data is detected and the first object information representing the position is output. Furthermore, the predicted position of the surrounding objects of the vehicle at a time that is a predetermined time later than the time of generation of the first data is detected and the first object prediction information representing the predicted position is output. The position of an object in the vicinity of the vehicle is detected from one of a plurality of second data representing the surrounding conditions of the vehicle, generated in a time sequence by a second sensor different from the first sensor mounted on the vehicle, and second object information representing that position is output. Predict whether the autonomous driving level controlling the vehicle's driving will change from a first autonomous driving level where the driver has a first monitoring duty to a second autonomous driving level where the driver has a second monitoring duty different from the first monitoring duty. If the prediction is that the vehicle will change, determine the time of the predicted change. as well as The first object information is used to create the driving path of the vehicle until the time of change is earlier than the predetermined time. The first object prediction information is used to create the driving path from the time of change to the time of change. After the time of change, the second object information is used to create the driving path.

Images