Action planning device, and coupled-vehicle automatic driving system

Abstract

A coupled-vehicle automatic driving system which has an action planning device (100), wherein: the action planning device (100) comprises an obstacle information acquisition unit (10) that acquires information about an obstacle around a coupled vehicle comprising a towing vehicle (VTr) which is an automatic driving vehicle and a towed vehicle (Tr) which is towed thereby, an action planning unit (50) that plans an action which includes automatic parking of the towing vehicle (VTr), and a parking region obstacle determination unit (40) that determines whether or not an obstacle is present in a region through which the coupled vehicle travels when automatic parking is performed in a parking space (PSt) set by the action planning unit (50); and the action planning unit (50) does not initiate automatic parking in the parking space (PSt) when the parking region obstacle determination unit (40) determines that an obstacle is present.

Description

Action Planning Device and Connected Vehicle Automatic Driving System

[0001] The present disclosure relates to an action planning device and a connected vehicle automatic driving system.

[0002] There is a distribution center as a facility serving as a base for transporting and delivering goods. For example, a truck, which is a connected vehicle composed of a tractor as a towing vehicle and a trailer towed by the tractor with goods loaded thereon, enters the distribution center. The trailer, which is the towed vehicle with goods loaded thereon, is parked within the premises of the distribution center. The goods loaded on the trailer are carried into a warehouse provided in the distribution center, and the trailer from which the goods have been loaded and unloaded is parked within the premises of the distribution center as an empty trailer. Also, the empty trailer is loaded with goods carried out from the warehouse. Thereafter, the trailer loaded with goods is towed by the tractor and carried out from the distribution center, whereby the goods are transported. In this way, the transportation of goods is carried out with the distribution center as a base.

[0003] In order to realize the transportation of goods with the distribution center as a base as described above, it is necessary to transport the trailer from the position where the trailer is parked to the warehouse provided in the distribution center, or to transport the trailer from the warehouse to the parking position. Conventionally, the transportation of the trailer has been carried out by an operator using a vehicle driven by the operator or a transport device such as a trailer dolly. On the other hand, it is required to automatically or semi-automatically transport the trailer using a transport vehicle such as an autonomous driving vehicle capable of autonomous driving.

[0004] When towing a trailer with an autonomous driving transport vehicle, it is necessary to avoid a collision by an automatic brake or the like when there is an obstacle that may cause a collision. For example, Patent Document 1 conventionally discloses an obstacle detection method around a connected vehicle considering a trailer when a transport vehicle and a trailer are connected and run as a connected vehicle.

[0005] In the technology disclosed in Patent Document 1, in a connected vehicle including a towing vehicle and a towed vehicle, when the towing vehicle turns, the passing areas of both are set as a monitoring area so that entrainment does not occur due to the difference in the turning areas of both, and an obstacle is detected using a sensor, particularly a side sensor, installed on the towing vehicle or the towed vehicle.

[0006] International Publication No. 2023 / 063224

[0007] However, in the transportation of trailers at a distribution center, trailer changes are necessary, which imposes a limitation for autonomous transport vehicles where the towing vehicle cannot have obstacle detection sensors installed on the trailer itself. In other words, it is not possible to use sensors installed on the towed trailer, as described in Patent Document 1.

[0008] Furthermore, when performing automatic parking by towing a trailer, there is a problem in that a blind spot occurs in the sensors installed on the towing vehicle when reversing. This blind spot becomes larger the longer the trailer is. However, Patent Document 1 only discloses obstacle detection and driving methods during turning, and does not disclose any specific measures to address the blind spot when reversing.

[0009] This disclosure discloses technology to solve the above-mentioned problems, and aims to provide a behavior planning device and an automated driving system for coupled vehicles that can determine the presence or absence of obstacles in the area in which coupled vehicles, including a towing vehicle and a towed vehicle, travel when they are automatically parking, and can avoid collisions with obstacles.

[0010] The action planning device of this disclosure includes: an obstacle information acquisition unit that acquires information on obstacles around a connected vehicle including a towing vehicle which is an autonomous driving vehicle and a towed vehicle towed by the towing vehicle; an action planning unit that plans actions including automatic parking of the towing vehicle; and a parking area obstacle determination unit that determines whether or not the obstacle exists in the area in which the connected vehicle travels when automatically parking in a parking space set by the action planning unit, wherein the action planning unit does not start the execution of automatic parking in the parking space if the parking area obstacle determination unit determines that an obstacle exists.

[0011] According to this disclosure, before entering a parking space, the system detects the presence or absence of obstacles within the parking space and along the parking path, which are the areas in which the connected vehicle will travel. If there are obstacles, the system will not perform automatic parking. In other words, it will only perform parking after confirming that there are no obstacles, thus making it possible to avoid collisions with obstacles.

[0012] This is a functional block diagram showing the configuration of the automated trailer vehicle system according to Embodiment 1. This is a diagram showing an example of the obstacle detection range by a vehicle sensor mounted on a towing vehicle. This is a diagram for explaining the operation of a trailer vehicle at a distribution center to which the automated trailer vehicle system according to Embodiment 1 is applied. This is a diagram for explaining the target parking space and target route when a trailer vehicle parks. This is a flowchart showing the operation of the automated trailer vehicle system according to Embodiment 1. This is a flowchart showing the operation of the automated trailer vehicle system according to Embodiment 1. This is a diagram showing an example of covering the blind spot area of ​​the vehicle sensor with a roadside sensor. This is a diagram showing an example of the hardware configuration of the automated trailer vehicle system according to Embodiment 1.

[0013] The following description of this embodiment will be given with reference to the drawings. In each drawing, the same reference numerals indicate the same or corresponding parts. The drawings are schematic, and for the sake of clarity, some components may be omitted or simplified as appropriate. Furthermore, the relative sizes and positions of components shown in different drawings are not necessarily accurately represented and may be changed as appropriate. In addition, hatching may be added to drawings such as plan views, which are not cross-sectional views, to facilitate understanding of the embodiment. In the following description, similar components will be shown with the same reference numerals, and their names and functions will also be the same. Therefore, detailed explanations of them may be omitted to avoid redundancy. Furthermore, in the following description, even if terms such as "top," "bottom," "left," "right," "side," "bottom," "front," or "back" that mean specific positions and directions are used, these terms are used for convenience to facilitate understanding of the embodiment and are not related to the actual direction in which it is implemented.

[0014] Embodiment 1. <Configuration of the Automated Twisted Vehicle System> Figure 1 is a functional block diagram showing the configuration of the automated trailer system according to Embodiment 1. As will be described in detail later, the trailer system here consists of a trailer on which goods are loaded and a tractor that tows it, and the vehicle performs automated driving. In the following description, when referred to as a trailer system, it includes the trailer which is the towed vehicle and the tractor which is the towing vehicle, and when referred to as a vehicle, it refers to the tractor which is the towing vehicle. In Figure 1, the automated trailer system 1000 includes a vehicle sensor 1 mounted on the vehicle that detects obstacle information, a roadside sensor 2 mounted on the roadside that detects obstacle information, an action planning device 100 that acquires obstacle information and plans the vehicle's actions, and a drive system control device 200 mounted on the vehicle that drives the vehicle.

[0015] The action planning device 100 includes an obstacle information acquisition unit 10 that acquires information on obstacles around the coupled vehicle, a road information acquisition unit 20 that acquires information on the roads around the coupled vehicle and the roads on which the coupled vehicle travels, a vehicle position acquisition unit 30 that acquires the position information of its own vehicle, a parking area obstacle determination unit 40 that determines whether or not there are obstacles in the area on which the coupled vehicle travels when parked, an action planning unit 50 that plans the vehicle's actions, and a vehicle target path generation unit 60.

[0016] The obstacle information acquisition unit 10 acquires information about obstacles present around the vehicle from the vehicle sensor 1 mounted on the vehicle and the roadside sensor 2 installed on the road, etc. Obstacles include, for example, other vehicles, pedestrians, bicycles, motorcycles and other moving objects present around the vehicle, and fixed obstacles such as containers and trailers placed on the road.

[0017] The vehicle sensor 1 is, for example, at least one of a camera, radar, LiDAR (Light Detection and Ranging), and sonar sensor (ultrasonic sensor) installed on the vehicle. The vehicle sensor 1 detects obstacles present around the vehicle, and may output the obstacle information to the obstacle information acquisition unit 10 as obstacle information linked to a classification type such as other vehicles, pedestrians, bicycles, or motorcycles.

[0018] The cameras are positioned to capture images of the front, sides, and rear of the vehicle. From the captured images, the system obtains information about the vehicle's environment, such as lane markings, parking space lines, and obstacles in front of the vehicle.

[0019] Radar measures the relative distance and relative speed of obstacles around a vehicle by emitting radio waves around the vehicle and detecting the reflected waves, and outputs the measurement results. A typical example is millimeter-wave radar.

[0020] LiDAR detects the position and distance of objects by shining a laser around a vehicle and measuring the time difference between the laser beam reflecting off surrounding objects and returning.

[0021] Sonar sensors detect the location and distance of objects by emitting ultrasonic waves around a vehicle and measuring the time difference between the time it takes for the waves to reflect back from surrounding objects.

[0022] Figure 2 shows an example of obstacle detection range by vehicle sensors in an example where multiple vehicle sensors described above are combined and placed on a tractor, which is a towing vehicle. First, the configuration of the tractor VTr will be briefly explained. Figure 2 is a top view of the tractor VTr, with a box-shaped section called the cab VTr_a containing the driver's seat at the front, a coupler VTr_b that connects to the kingpin of the trailer to be towed at the rear near the rear wheels, and mechanical components such as the engine mounted on the frame at the bottom.

[0023] In Figure 2, four vehicle sensors 1, from vehicle sensor 1a to vehicle sensor 1d, are mounted on the tractor VTr vehicle. One vehicle sensor 1a is positioned at the front, two vehicle sensors 1b and 1c are positioned to the sides, and one vehicle sensor 1d is positioned below the rear coupler Vtr_b. The detection range of each vehicle sensor 1 is shown by a dashed fan shape. The tractor VTr can detect obstacles in the areas of detection range SN1 in front, detection ranges SN2 and SN3 to the sides, and detection range SN4 to the rear. Note that there may be blind spots behind the trailer due to attached structures or tires. Also, the number of vehicle sensors 1 is not limited to four, but it is desirable to be able to set the obstacle detection range to cover the area around the tractor VTr.

[0024] The roadside sensor 2 is installed, for example, on the roadside and is equipped with at least one of a camera, radar, or LiDAR. The obstacle information acquisition unit 10 acquires information about the area around the roadside detected by the roadside sensor 2 as obstacle information via wireless communication. Note that the roadside sensor 2 may be installed not only on the roadside but also on the roadway, shoulder, sidewalk, etc., or on buildings, utility poles, etc., near the road. Furthermore, information about the area around the roadside may be acquired from a remote control system, etc. In addition, if the purpose is to detect obstacles in a specific area where articulated vehicles travel, for example, a distribution center, the sensors may be placed around the distribution center or near the warehouse. As will be described later, if a control center is located within the distribution center, obstacle information may be acquired from the control center.

[0025] The road information acquisition unit 20 acquires information about roads around the vehicle. The road information acquisition unit 20 includes a map information acquisition unit 22 that has previously acquired map data of the vehicle's planned route, and acquires road information around the vehicle based on the vehicle position information acquired by the vehicle position acquisition unit described later. It may also include a road information detection unit (not shown) other than the map information acquisition unit 22.

[0026] Here, the map data includes road information consisting of the centerline of the lane the vehicle travels in, lane width information, intersection stop line information, pedestrian crossing information, information on the number of branches at intersections such as T-junctions and crossroads, and information on the start position of lane splitting. In addition, the road information included in the map data may be obtained from the surrounding structure detection results obtained from at least one of the vehicle sensors 1 equipped on the vehicle, which include a camera, radar, LiDAR, and sonar sensor.

[0027] Furthermore, if the movement of the articulated vehicles is within the distribution center, it is sufficient to obtain map information for an area that includes the distribution center and the flow of vehicles to and from its entrances and exits. This information may be obtained from the surrounding structure detection results obtained from the vehicle sensor 1 and roadside sensor 2 installed on the vehicle, or it may be map information that includes obstacle information obtained from the vehicle sensor 1 and roadside sensor 2.

[0028] Furthermore, the obstacle information acquisition unit 10 calculates the sensor blind spot area in which the vehicle sensor 1 and roadside sensor 2 cannot detect obstacles, based on information about obstacles present around the vehicle acquired from the vehicle sensor 1 mounted on the vehicle and the roadside sensor 2 installed on the road, etc., and map information acquired from the road information acquisition unit 20.

[0029] The vehicle position acquisition unit 30 is equipped with a GNSS (Global Navigation Satellite System) sensor to determine the position of its own vehicle. A GNSS antenna is connected to the GNSS sensor, and the GNSS antenna receives positioning signals from positioning satellites orbiting in the satellite orbit. The received positioning signals are analyzed, and information about the phase center of the GNSS antenna (latitude, longitude, altitude, and azimuth, etc.) is output. In this way, the vehicle position acquisition unit 30 acquires the position information of its own vehicle.

[0030] Furthermore, in addition to the method using GNSS sensors, the vehicle position may also be acquired using SLAM (Simultaneous Localization and Mapping) technology, which utilizes the surrounding structure detection results obtained from at least one of the following: cameras, radar, LiDAR, and sonar sensors. If the vehicle is not equipped with a behavior planning device 100, the vehicle position acquisition unit 30 may acquire vehicle position information from the GNSS sensors mounted on the vehicle via communication.

[0031] The target route generation unit 60 generates target route information, including a target route for the connected vehicle to park, based on the target parking space set by the action planning unit 50, and information obtained from the obstacle information acquisition unit 10, the road information acquisition unit 20, and the vehicle position acquisition unit 30. The target route information includes information such as the target position and target vehicle speed of the vehicle.

[0032] The parking area obstacle determination unit 40 determines whether there are obstacles in the area where the connected vehicle travels when parking, and whether there are sensor blind spots, based on the target path information generated by the target path generation unit 60, and the information obtained from the obstacle information acquisition unit 10, the road information acquisition unit 20, and the vehicle position acquisition unit 30. Here, the area where the connected vehicle travels refers to the area near the path, which includes the path the connected vehicle travels and a predetermined distance from that path. The area where the connected vehicle travels when parking refers to the area near the path, which includes the path the connected vehicle travels to the parking space, and a predetermined distance from that path, as well as the area within the parking space. Sensor blind spots will be described later.

[0033] The drive system control device 200 controls actuators such as the steering, brakes, and accelerator to follow the target path generated by the target path generation unit 60. This controls the vehicle's behavior, i.e., its lateral and longitudinal movements, causing the vehicle to travel along the target path.

[0034] <Distribution Center Situation> Figure 3 is a diagram illustrating an example of the operation of articulated vehicles in a distribution center where the articulated vehicle automated driving system 1000 is applied. In Figure 3, tractor VTr1, which is towing trailer Tr1, parks in parking space PS and then disengages from trailer Tr1. Multiple trailers Tr, including trailer Tr6, are parked in parking space PS. Tractor VTr2 is towing trailer Tr2, which is loaded with goods, to the receiving space RS for bringing the goods into the warehouse. Tractor VTr3 is in the process of coupling to tow trailer Tr3, which has been loaded with goods from the warehouse, to parking space PS. Tractor VTr4 is towing trailer Tr4, which has been unloaded from goods in the warehouse, to parking space PS. Tractor VTr5 has towed trailer Tr5, which is loaded with goods, to the receiving space RS, then disengaged and is heading towards parking space PS to tow another trailer. Trailer Tr7 is connected to the loading space RS, and goods have been loaded and unloaded. At this point, each tractor VTr operates automatically.

[0035] By repeating this process, the transportation of goods from the parking space PS to the receiving space RS within the distribution center is automated. Note that articulated vehicles arriving from outside the distribution center may proceed directly to the receiving space RS without passing through the parking space PS. Furthermore, roadside sensors 2 are installed in the distribution center to monitor the operating areas of the articulated vehicles and tractor VTr, as well as the conditions of the parking space PS and receiving space RS, and to detect obstacles. Since the trailer Tr is longer and taller than the tractor VTr, blind spots are likely to occur near where the trailer Tr is parked. Therefore, it is desirable to eliminate blind spots by installing multiple roadside sensors 2, or by positioning them to cover the entire distribution center from a high vantage point. The placement of roadside sensors 2 will be described later.

[0036] This example also shows a distribution center with a control center 300 installed. The control center 300 performs functions such as detecting obstacles within the distribution center, similar to the roadside sensors 2, acquiring obstacle information detected by the roadside sensors 2 within the distribution center, acquiring obstacle information detected by the vehicle sensors 1 of tractors VTr traveling within the distribution center, and acquiring the status of trailers Tr parked in parking spaces PS and receiving spaces RS within the distribution center. In other words, the control center 300 has obstacle information and parking space information. The control center 300 also performs functions such as assigning parking spaces to each trailer Tr or to the action planning device 100, and providing obstacle information. Note that the control center 300 is not required.

[0037] <Operation of the Automated Trolley Vehicle System> Next, the operation of the automated trailer vehicle system according to Embodiment 1 will be explained using Figures 4 and 5A, 5B. Figure 4 is a diagram illustrating the target parking space and target route when the trailer vehicle is parked, and Figures 5A and 5B are flowcharts illustrating the operation of the automated trailer vehicle system. The operations in Figures 5A and 5B are repeatedly performed while the trailer vehicle is in motion. Each step in Figures 5A and 5B will be explained in correspondence with the functional parts shown in the functional block diagram of Figure 1. In the following explanation, an example will be described in which the trailer vehicle parks in a distribution center in an area sandwiched between trailers Tr, such as between trailers Tr6 and Tr7 in the parking space PS in Figure 3, or between trailers Tr5 and Tr3 in the receiving space RS.

[0038] First, in step S101, the obstacle information acquisition unit 10 acquires information such as the position, speed, size, and type of obstacles present around the connected vehicle as detected by the vehicle sensor 1 and the roadside sensor 2.

[0039] In step S102, the vehicle position acquisition unit 30 acquires the position of its own vehicle, specifically the position of the tractor VTr of the connected vehicle. In step S103, the road information acquisition unit 20 acquires map information of the distribution center as road information.

[0040] Based on the information obtained in step S101 and step S103, the obstacle information acquisition unit 10 calculates a sensor dead zone area, which is an area where neither the vehicle sensor 1 nor the roadside sensor 2 can detect an obstacle, that is, an area where both sensors cannot detect an obstacle.

[0041] In step S104, the action planning unit 50 calculates the target parking frame PSt shown in FIG. 4. Here, the target parking frame PSt is calculated and set based on an instruction from the control center 300 of the distribution center or information on the vacant parking area detected by the roadside sensor 2. Here, the target parking frame PSt is an area sandwiched between the parked trailers Tr11 and Tr12.

[0042] In step S105, based on the obstacle information acquired by the obstacle information acquisition unit 10, the map information of the distribution center acquired by the road information acquisition unit 20, and the position information of the tractor VTr0 acquired by the vehicle position acquisition unit 30, the target route generation unit 60 calculates and generates a target route Rot of the tractor VTr0 for the connected vehicle to park with respect to the target parking frame PSt set by the action planning unit 50.

[0043] In step S106, the parking area obstacle determination unit 40 determines whether there is a sensor dead zone area in the target route Rot, its vicinity, and within the target parking frame PSt, which is the area where the connected vehicle travels. If there is a sensor dead zone area (Yes in step S106), the process proceeds to step S109; if not (No in step S106), the process proceeds to step S107.

[0044] In step S107, the parking area obstacle determination unit 40 determines whether there is an obstacle in the area where the connected vehicle travels according to the target route Rot. If there is an obstacle (Yes in step S107), the process proceeds to step S111; if there is no obstacle (No in step S107), the process proceeds to step S108. [[ID=I3]]

[0045] In step S108, the parking area obstacle determination unit 40 determines whether there is an obstacle within the target parking frame PSt. If there is an obstacle (Yes in step S108), the process proceeds to step S109. If there is no obstacle (No in step S108), the process proceeds to step S112.

[0046] In step S109, the action plan unit 50 determines whether there is a candidate for a target parking frame different from the current target parking frame PSt. If there is a candidate for a different target parking frame (Yes in step S109), the process proceeds to step S110. If there is no candidate for a different target parking frame (No in step S109), the process proceeds to step S111.

[0047] In step S110, the action plan unit 50 selects from the candidates for a different target parking frame and changes the target parking frame PSt. The selection method may be a method based on a predetermined priority order, a method of selecting the candidate closest to the current target parking frame PSt, a method of selecting a candidate with no blind spot detected by the roadside sensor 2 and no obstacle detected, etc., and any method may be used. When the target parking frame PSt is changed, the process returns to step S105.

[0048] In step S111, the action plan unit 50 determines not to execute automatic parking, and the articulated vehicle waits on the spot. In step S112, the action plan unit 50 determines to start the execution of automatic parking, and instructs the target route generation unit 60 and the drive system control device 200 of the tractor VTr0 of the articulated vehicle to execute automatic parking. The tractor VTr0 starts the operation of automatic parking to the target parking frame along the target route.

[0049] Note that it may take some time before automatic parking can be performed, such as by repeatedly performing the operation to change the target parking space, returning from step S110 to step S105. Therefore, in step S113, it is determined whether a preset time has elapsed since acquiring obstacle detection information for the target route Rot, its vicinity, and within the target parking space Pst. If the preset time has elapsed (Yes in step S113), the system proceeds to step S114; if the preset time has not elapsed (No in step S113), the system proceeds to step S115.

[0050] In step S114, the action planning unit 50 determines not to perform automatic parking, that is, to interrupt automatic parking, and the coupled vehicle waits in place. Then the process returns to step S101. In step S115, the action planning unit 50 determines to continue automatic parking, and the coupled vehicle's tractor VTr0 continues the automatic parking operation.

[0051] <Eliminating Sensor Blind Spots> The obstacle information acquisition unit 10 calculates the sensor blind spot area based on the information obtained in steps S101 and S103. However, since the vehicle sensor 1 and the roadside sensor 2 are installed in different locations, the blind spots of each sensor can be covered by using information from obstacle information detection means (sensors) with different detection ranges.

[0052] <Example of Solution 1> For example, Figure 6 shows an example in which the vehicle sensor 1's blind spot area, where obstacles cannot be detected, is covered by the roadside sensor 2. When the tractor VTr0 parks in the target parking space PSst according to the target path shown in Figure 4, area X (the area with dot hatching) becomes a blind spot area for the vehicle sensor 1 at the current position of the coupled vehicle. However, as shown in Figure 6, if the roadside sensor 2 is installed behind the target parking space PSst, area X can be largely covered by the detection range SNrsu of the roadside sensor 2.

[0053] Thus, blind spots for the vehicle sensor 1 are created by the trailer Tr towed by the tractor VTr itself and by the parked trailer Tr. However, by installing roadside sensors 2 in positions where such blind spots are expected to occur, the blind spots of the vehicle sensor 1 can be covered by the roadside sensors 2. Specifically, it is desirable to install multiple roadside sensors 2 on the warehouse side, which is the back of the parking space PS, or on the warehouse side, which is the back of the loading space RS.

[0054] <Example of Solution 2> Conversely, in the roadside sensor 2 blind spot area where a blind spot is expected to occur, the vehicle can be driven in such a way that the vehicle sensor 1 scans the roadside sensor 2 blind spot area, thereby covering the roadside sensor 2 blind spot area with the vehicle sensor 1. For example, when tractor VTr0 parks in target parking space PSt according to the target path shown in Figure 4, starting from point P0, tractor VTr0 travels towards target parking space PSt to point P1, so vehicle sensors 1a and 1b can detect obstacles near the parking entrance of target parking space PSt. While tractor VTr0 travels from point P1 to point P2, vehicle sensors 1b and 1d can detect obstacles near the parking entrance of target parking space PSt. When tractor VTr0 proceeds to point P3, it moves in reverse towards point P4, turns around at point P4 and proceeds to target parking space PSt, and near point P4, vehicle sensor 1b can detect obstacles near the parking entrance of target parking space PSt. Thus, in Figure 6, in the area X near the parking entrance of the target parking space PSt, where the detection range SNrsu of the roadside sensor 2 does not cover, the tractor VTr0 can cover the blind spot area of ​​the roadside sensor 2 by driving around the target parking space PSt.

[0055] In this way, the obstacle information acquisition unit 10 pre-calculates the roadside sensor blind spot area where the roadside sensor 2 cannot detect obstacles, based on the information of obstacles present around the connected vehicle acquired from the roadside sensor 2 and the map information acquired from the road information acquisition unit 20. The action planning unit 50 can cover the roadside sensor blind spot area by planning the tractor VTr's travel path so that the vehicle sensor 1 scans the roadside sensor blind spot area of ​​the roadside sensor 2. In particular, if there is a roadside sensor blind spot area in or near a parking space PS where a parking space is set, it is effective to plan the tractor VTr's travel so that the vehicle sensor 1 scans that roadside sensor blind spot area, as this area is the area where the connected vehicle travels, thereby covering the roadside sensor blind spot area. Furthermore, by storing the travel path of the tractor VTr equipped with the vehicle sensor 1 and obstacle information on the map information over time, the roadside sensor blind spot area can be covered.

[0056] <Example of Solution 3> When a distribution center is equipped with a control center 300, the control center 300 understands the placement of the roadside sensors 2 and the blind spots of the roadside sensors 2, and also acquires obstacle information from the vehicle sensors 1 of each tractor VTr as it goes. Therefore, when a connected vehicle parks, the control center 300 distributes information about obstacles around the parking space to the tractor VTr of that connected vehicle or to the action planning device 100. This makes it possible to generate a target route that eliminates blind spots, and to select a route and parking space without obstacles, thus enabling smooth parking.

[0057] Figure 7 shows an example of the hardware configuration of the action planning device 100, control center 300, and articulated vehicle automatic driving system 1000 according to Embodiment 1. The action planning device 100, control center 300, and articulated vehicle automatic driving system 1000 may include a calculation processing circuit 1100, a storage device 1200 including a read-only memory (ROM) that stores programs for executing the functions of each functional unit, and random access memory (RAM) that stores the execution results data of each functional unit which are the calculation results of the program, as well as the acquired data, an input / output circuit 1300, and a communication circuit 1400.

[0058] The arithmetic processing circuit 1100 includes a processor configured using a CPU (Central Processing Unit), which may be a digital signal processor (DSP) or a logic circuit. Dedicated hardware may be applied to the arithmetic processing circuit 1100. When the arithmetic processing circuit 1100 is dedicated hardware, it may be, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a combination thereof. Furthermore, each component and the functional parts they possess can realize their functions using arithmetic processing circuits as dedicated hardware, while other functions can be realized using software. In other words, the aforementioned functions can be realized through hardware, software, or a combination thereof.

[0059] The communication circuit 1400 can use a communication module that conforms to, for example, LTE (Long Term Evolution), 4G (4th Generation), or 5G (5th Generation).

[0060] The action planning device 100 may be mounted on the tractor VTr vehicle, or it may exist separately from the vehicle. In addition, if the distribution center has a control center 300, the action planning device 100 may be installed in the control center 300, or the control center 300 may have the functions of the action planning device 100.

[0061] At least roadside sensor 2 transmits obstacle information to action planning device 100 via communication. When action planning device 100 is mounted on a vehicle, for example, communication between vehicle sensor 1, action planning device 100, and drive system control device 200 is performed using the in-vehicle communication protocol CAN (Controller Area Network; registered trademark).

[0062] <Regarding Modifications of Embodiments> In the embodiments described above, the dimensions, shapes, relative arrangements, or conditions of implementation of each component may also be described, but these are all examples and are not limited to those described in this specification. Accordingly, countless modifications and equivalents not shown are conceivable within the scope of the art disclosed in this specification. For example, this includes modifying, adding, or omitting at least one component, or even extracting at least one component from at least one embodiment and combining it with a component from another embodiment.

[0063] Furthermore, unless contradictory, a component described as being provided as "one" in the embodiments described above may be provided as "one or more." In addition, each component in the embodiments described above is a conceptual unit, and the scope of the technology disclosed in this specification includes cases where one component consists of multiple structures, where one component corresponds to a part of a structure, and where multiple components are provided in a single structure. Furthermore, each component in the embodiments described above includes structures having other structures or shapes, as long as they perform the same function. Also, the descriptions in this specification are referenced for all purposes relating to the technology and none of them are considered prior art.

[0064] As described above, the action planning device of this embodiment 1 includes an obstacle information acquisition unit that acquires information on obstacles around the connected vehicle, including the towing vehicle which is an autonomous vehicle and the towed vehicle towed by it; an action planning unit that plans actions including the automatic parking of the towing vehicle; and a parking area obstacle determination unit that determines whether or not there are obstacles in the area in which the connected vehicle travels when it automatically parks in a parking space set by the action planning unit. If the parking area obstacle determination unit determines that there are obstacles, the action planning unit will not start the execution of automatic parking in the parking space, that is, it will perform parking after confirming that there are no obstacles, so it is possible to avoid collisions with obstacles.

[0065] Furthermore, the obstacle information acquisition unit acquires obstacle information around the towed vehicle by using vehicle sensors and roadside sensors mounted on the towing vehicle, that is, by combining obstacle detection by vehicle sensors and obstacle detection by roadside sensors. This allows the roadside sensors to cover blind spots covered by the vehicle sensors. Therefore, when the towed vehicle is reversing, even if there is an obstacle in the blind spot of the vehicle sensors caused by the towed trailer, the roadside sensors can cover it, making it possible to avoid collisions with obstacles.

[0066] Furthermore, if the parking area obstacle detection unit determines that an obstacle exists within a parking space, the action planning unit plans to automatically park in a different parking space. This improves the tractor's operational efficiency by eliminating the need to wait for the obstacle to disappear.

[0067] Furthermore, the obstacle information acquisition unit calculates the roadside sensor blind spot area where obstacles cannot be detected by the roadside sensor, and the action planning unit plans the towing vehicle's movement so that the vehicle sensor can scan the roadside sensor blind spot area. This makes it possible to cover the sensor blind spot area of ​​the roadside sensor with the vehicle sensor. In particular, when the area in which the towed vehicle travels overlaps with the roadside sensor blind spot area, for example near a parking space, smooth automatic parking becomes possible by covering the sensor blind spot area of ​​the roadside sensor with the vehicle sensor.

[0068] Furthermore, if the parking area obstacle detection unit determines that no obstacles exist, the action planning unit will perform automatic parking if a preset time has not elapsed since the obstacle information acquisition unit obtained information that no obstacles exist. If the preset time has elapsed, the unit will not perform automatic parking, will instead obtain obstacle information again from the obstacle information acquisition unit, and the parking area obstacle detection unit will determine whether or not there are obstacles in the area where the coupled vehicle will travel. This makes it possible to respond even if the presence of obstacles changes over time. In other words, if a new obstacle is detected over time, a collision can be avoided.

[0069] Furthermore, the obstacle information acquisition unit calculates sensor blind spots where neither the vehicle sensor nor the roadside sensor can detect obstacles, and the action planning unit, when the coupled vehicle is to automatically park in a parking space set by the action planning unit, will not start the automatic parking process if the parking area obstacle determination unit determines that a sensor blind spot exists in the area where the coupled vehicle will travel. This reduces the risk of detection failure and avoids collisions with obstacles if they are present in the sensor blind spots.

[0070] The system further includes a target route generation unit that generates target route information, including the target route of the towing vehicle to the parking space set by the action planning unit, and a parking area obstacle determination unit that determines whether or not there is an obstacle in the area in which the towed vehicle is traveling when the towed vehicle is traveling according to the target route. By including the route to the parking space in the area to be determined for the presence or absence of obstacles, the possibility of avoiding collisions with obstacles can be further increased.

[0071] Furthermore, according to the linked vehicle automatic driving system of this embodiment 1, the system includes a linked vehicle including a towing vehicle and a towed vehicle, vehicle sensors, roadside sensors, and the aforementioned action planning device, and the towing vehicle is made to drive automatically according to the target route information calculated by the action planning device. As a result, the sensor blind spots in the area in which the linked vehicle is traveling can be eliminated as much as possible, and parking is performed after confirming that there are no obstacles, making it possible to avoid collisions with obstacles.

[0072] Furthermore, in the above-mentioned automated articulated vehicle driving system, the articulated vehicle travels through the distribution center, and the action planning device plans actions, including the automatic parking of the towing vehicle, based on obstacle information and parking space information held by the control center located at the distribution center. This makes it possible to avoid collisions with obstacles within the distribution center.

[0073] While the first embodiment uses automated parking at a distribution center as an example, it is not limited to this. For example, even when a combined vehicle traveling between distribution centers parks in a parking area, it goes without saying that if the tractor of the combined vehicle is equipped with a behavior planning device and vehicle sensors, and information is acquired from roadside sensors and control systems, automated parking or parking assistance that avoids collisions with obstacles is possible. Furthermore, it can also be applied in situations such as parking a trailer at a port or loading a trailer onto a car ferry, as long as sensor information that can replace roadside sensors is available.

[0074] While this disclosure describes exemplary embodiments, the various features, aspects, and functions described in the embodiments are not limited to the application of any particular embodiment, but can be applied individually or in various combinations to the embodiments. Accordingly, countless variations not illustrated are conceivable within the scope of the art disclosed in this specification. These include, for example, modifications, additions, or omissions of at least one component.

[0075] 1, 1a, 1b, 1c, 1d: Vehicle sensors, 2: Roadside sensors, 10: Obstacle information acquisition unit, 20: Road information acquisition unit, 22: Map information acquisition unit, 30: Vehicle position acquisition unit, 40: Parking area obstacle determination unit, 50: Action planning unit, 60: Target route generation unit, 100: Action planning device, 200: Drive system control device, 300: Control center, 1000: Connected vehicle automatic driving system, 1100: Calculation processing circuit, 1200: Memory device, 1300: Input / output circuit, 1400: Communication circuit, VTr: Tractor, Tr: Trailer, SN1, SN2, SN3, SN4: Detection range, PS: Parking space, PSt: Target parking space, RS: Load receiving space, ROT: Target route, SNrsu: Detection range, X: Area.

Claims

1. An action planning device comprising: an obstacle information acquisition unit that acquires information on obstacles around a connected vehicle including a towing vehicle which is an autonomous vehicle and a towed vehicle towed by the towing vehicle; an action planning unit that plans actions including automatic parking of the towing vehicle; and a parking area obstacle determination unit that determines whether or not the obstacle exists in the area in which the connected vehicle travels when automatically parking in a parking space set by the action planning unit, wherein the action planning unit does not start the execution of automatic parking in the parking space if the parking area obstacle determination unit determines that an obstacle exists.

2. The action planning device according to claim 1, wherein the obstacle information acquisition unit acquires obstacle information around the coupled vehicle using a vehicle sensor and a roadside sensor mounted on the towing vehicle.

3. The action planning device according to claim 1 or 2, wherein the action planning unit plans automatic parking to another parking space if the parking area obstacle determination unit determines that an obstacle exists within the parking space.

4. The action planning device according to claim 2, wherein the obstacle information acquisition unit calculates the roadside sensor blind spot area in which the roadside sensor cannot detect the obstacle, and the action planning unit plans the movement of the towing vehicle so that the vehicle sensor can scan the roadside sensor blind spot area.

5. The action planning device according to any one of claims 1 to 4, wherein the action planning unit, when the parking area obstacle determination unit determines that no obstacle exists, performs automatic parking if a preset time has not elapsed since the obstacle information acquisition unit obtained information that no obstacle exists, and if a preset time has elapsed, the obstacle information acquisition unit obtains information about the obstacle again, and the parking area obstacle determination unit determines whether or not the obstacle exists in the area where the connected vehicle will travel.

6. The action planning device according to claim 2 or 4, wherein the obstacle information acquisition unit calculates a sensor blind spot area in which neither the vehicle sensor nor the roadside sensor can detect the obstacle, and the action planning unit does not start the execution of automatic parking in the parking space if the parking area obstacle determination unit determines that a sensor blind spot area exists in the area in which the connected vehicle is traveling when the connected vehicle is to be automatically parked in the parking space set by the action planning unit.

7. The action planning device according to any one of claims 1 to 6, further comprising a target path generation unit that generates target path information including a target path for the towing vehicle to the parking space set by the action planning unit, wherein the parking area obstacle determination unit determines whether or not the obstacle exists in the area through which the towed vehicle travels when the towed vehicle travels according to the target path.

8. An automated driving system for coupled vehicles, comprising the coupled vehicle including the towing vehicle and the towed vehicle, a vehicle sensor, a roadside sensor, and an action planning device according to any one of claims 1 to 7, wherein the system causes the towing vehicle to be driven automatically according to an action plan planned by the action planning device.

9. The linked vehicle travels through a distribution center, and the action planning device plans actions, including the automatic parking of the towing vehicle, based on obstacle information and parking space information held by a control center located at the distribution center, according to claim 8.

Images