Vehicle automatic driving method, automatic driving platform and vehicle system

Abstract

Disclosed are a method, platform, and vehicle system for autonomous driving of a vehicle. The method includes determining at least one of the motion status of an object detected by a sensor, its positional relationship with a reference object, and the distribution status of the object as obstacle-related information. Based on the obstacle-related information, a determination can be made as to whether the object is an obstacle. If the object is determined not to be an obstacle, the vehicle's driving direction remains unchanged. This allows accurate distinction between various other objects and obstacles, thereby avoiding unnecessary vehicle driving changes due to "false" obstacles.

Description

Technical Field

[0001] The present invention relates to vehicle control. Background Art

[0002] Research into autonomous driving has been actively pursued. For example, Japanese Unexamined Patent Application Publication No. 2018-132015 (JP 2018-132015A) discloses a vehicle system in which, in addition to an engine ECU, an autonomous driving ECU with sensing capabilities for detecting the vehicle's surroundings is also installed. The autonomous driving ECU issues commands to the engine ECU via an onboard network. The invention described in JP 2018-132015A allows for the addition of autonomous driving functionality to a vehicle without significantly altering the existing vehicle platform by providing separate ECUs for managing vehicle power and autonomous driving ECUs. Furthermore, third-party development of autonomous driving functionality is expected. Accurately identifying and determining obstacles is crucial in autonomous driving. Mistaking other objects for obstacles can cause the vehicle to undergo unnecessary maneuvers such as evasive maneuvers or deceleration. However, current autonomous driving methods and autonomous driving ECUs lack in-depth analysis of obstacle identification and determination, and obstacle identification is also insufficiently accurate. Summary of the Invention

[0003] The present invention has been made in view of the above-mentioned problems, and the present invention aims to provide an automatic driving method, an automatic driving platform and a vehicle system for a vehicle so as to accurately distinguish various other objects from obstacles, thereby avoiding the vehicle from undergoing unnecessary changes in driving operations due to "pseudo" obstacles.

[0004] According to a first aspect of the present disclosure, a method for automatic driving of a vehicle is provided. The method may include determining at least one of the motion state of an object detected by a sensor, its positional relationship with a reference object, and its distribution state as obstacle-related information. Based on the obstacle-related information, it may be determined whether the object is an obstacle. In the case where it is determined that the object is not an obstacle, the driving of the vehicle may not be changed. Objects that are easily misjudged as obstacles are significantly different from obstacles in terms of motion state, positional relationship with a reference object, and distribution state of the object. By taking into account at least one of the motion state of the object detected by the sensor, its positional relationship with a reference object, and distribution state of the object, "pseudo" obstacles can be accurately and quickly distinguished from obstacles, significantly reducing the error rate of obstacle judgment, thereby effectively avoiding the vehicle from undergoing unnecessary driving changes due to misjudgment of obstacles, thereby improving the experience of automatic driving.

[0005] In some embodiments, the object's motion condition may include the object's floating condition relative to the detection space, and the positional relationship between the object and the reference object may include the positional relationship between the object and the ground, whereby both together serve as obstacle-related information. In some embodiments, the object's motion condition may include the object's motion vector relative to the detection space, whereby both together serve as obstacle-related information. In some embodiments, the positional relationship between the object and the reference object may include the positional relationship between the object and the lane, and the object's distribution condition may include the continuous distribution of the object's reflection intensity, whereby both together serve as obstacle-related information.

[0006] According to a second aspect of the present disclosure, an autonomous driving platform is provided. The autonomous driving platform may be developed by a third party and include a sensor group and a first control unit that performs autonomous driving control of the vehicle. The sensor group may be configured to detect objects around the vehicle. The first control unit may be configured to determine, based on the information detected by the sensor group, at least one of the motion status of the detected object, the positional relationship with a reference object, and the distribution status of the object as obstacle-related information. The first control unit may be further configured to determine whether the object is an obstacle based on the obstacle-related information; and, if it is determined that the object is not an obstacle, issue a first control instruction that does not change the driving of the vehicle. The first control instruction may be converted into a second control instruction that can be interpreted by the vehicle platform via a vehicle control interface connecting the autonomous driving platform and the vehicle platform, so as to prevent unnecessary changes to the vehicle's driving due to misjudgment of obstacles via the vehicle platform.

[0007] According to a third aspect of the present disclosure, a vehicle system is provided for use in conjunction with a vehicle platform, wherein the vehicle platform includes a second control unit that performs driving control of the vehicle. The vehicle system may include an autonomous driving platform according to various embodiments of the present disclosure and the above-mentioned vehicle control interface. The vehicle control interface may include a third control unit, which may be configured to perform: obtaining the first control instruction from the first control unit; converting the first control instruction into a second control instruction for the second control unit; and sending the second control instruction to the second control unit. The vehicle system can cooperate with the vehicle platform through the autonomous driving platform and the vehicle control interface, and prevent unnecessary changes to the vehicle's driving due to misjudgment of obstacles via the vehicle platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Features, advantages, and technical and industrial significance of exemplary embodiments of the present invention will be described below with reference to the accompanying drawings, wherein like reference numerals represent like elements, and wherein:

[0009] Figure 1is a schematic diagram of a vehicle system according to a first embodiment;

[0010] Figure 2 is a block diagram schematically showing one example of components provided in the system;

[0011] Figure 3 is a diagram showing data input and output of a vehicle control interface;

[0012] Figure 4 is a diagram showing data to be converted;

[0013] Figure 5A is a flowchart showing the processing performed in the first embodiment;

[0014] Figure 5B is a flowchart showing the processing of the automatic driving method of a vehicle according to the second embodiment;

[0015] Figure 6A is a flowchart showing the processing performed in the first embodiment;

[0016] Figure 6B is a flowchart showing the processing of the automatic driving method of a vehicle according to the second embodiment;

[0017] Figure 7 is a diagram showing a vehicle travel plan;

[0018] Figure 8A A diagram showing physical control quantities (acceleration or deceleration) of the vehicle;

[0019] Figure 8B is a diagram showing a physical control amount (steering angle) of a vehicle; and

[0020] Figure 8C It is a graph showing the value of the physical control variable (acceleration or deceleration) of the vehicle at each time step. DETAILED DESCRIPTION

[0021] A configuration has been proposed in which a vehicle platform, including computers that control vehicle power, is independently provided from an autonomous driving platform that makes decisions about autonomous driving, with both platforms installed within the vehicle system. For example, the autonomous driving platform senses the vehicle's surroundings and, based on these sensing results, sends control instructions to the existing vehicle platform. This configuration allows independent vendors to develop their own platforms, thereby facilitating the development of autonomous driving functions by third parties.

[0022] At the same time, various issues arise when platforms developed by different suppliers are installed in the same vehicle system (i.e., the vehicle's powertrain and the autonomous driving system that issues control commands to that powertrain are connected to the same in-vehicle network). One of the issues that may arise is that the commands used to control the vehicle platform vary depending on the manufacturer and vehicle type. For example, the inputs or outputs of the engine ECU vary depending on the manufacturer and vehicle type, so designing an autonomous driving ECU compatible with all vehicle types is expensive. Furthermore, since various information used to control the vehicle flows through the in-vehicle network, it is undesirable to allow the autonomous driving platform (manufactured by a third party with no direct relationship with the vehicle platform) to have unrestricted access to that information.

[0023] Therefore, the vehicle system according to the present embodiment is configured such that the vehicle platform and the autonomous driving platform are connected via the vehicle control interface to relay information. Figure 1 is a schematic diagram of a vehicle system according to this embodiment. Vehicle platform 100 is a platform that includes a first computer (e.g., an engine ECU, etc., hereinafter also referred to as a "second control unit") that performs driving control of the vehicle. Autonomous driving platform 200 is a platform that includes a second computer (e.g., an autonomous driving ECU, hereinafter also referred to as a "first control unit") that performs autonomous driving control of the vehicle. Autonomous driving platform 200 may include a device for sensing the vehicle's surroundings and a device for generating a driving plan based on the sensing results. Figure 1 The vehicle system in the description includes the autonomous driving platform 200, the vehicle control interface 300, and the vehicle platform 100, but this is merely an example. In some embodiments, any two of the autonomous driving platform 200, the vehicle control interface 300, and the vehicle platform 100 may constitute the vehicle system. For example, a vehicle system including both the autonomous driving platform 200 and the vehicle control interface 300 may be provided. For example, the autonomous driving platform 200 and the vehicle control interface 300 may be manufactured and produced by a single supplier / third party and packaged as a kit or system.

[0024] The vehicle control interface 300 connects the vehicle platform 100 and the autonomous driving platform 200 and relays information between the two platforms. Specifically, the vehicle control interface 300 is configured to include a third control unit that retrieves a first control instruction from a second computer. The first control instruction is data for controlling the vehicle platform and includes at least one of first data related to a specified acceleration / deceleration and second data related to a specified driving trajectory. The third control unit converts the first control instruction into a second control instruction for the first computer and transmits the second control instruction to the first computer.

[0025] The first data is data related to the acceleration and deceleration of a specified vehicle, and the second data is data related to the specified driving trajectory. The first data may, for example, specify the speed change per unit time (acceleration or deceleration) or the target speed. Alternatively, the second data may specify the steering angle. The second data may also specify the driving trajectory.

[0026] The first control command is generated as a general command not specific to a first computer provided in the vehicle. The control unit converts the first control command into a second control command, which is data specific to the first computer. This configuration allows the general command to be converted into a command specific to a vehicle type or manufacturer.

[0027] If the first control instruction includes data other than the first data and the second data, the control unit may discard the data without converting the data. According to such a configuration, if data that should not be sent to the vehicle platform 100 is sent (for example, instructions for vehicle components that should not be accessed by the autonomous driving platform), such data can be appropriately filtered.

[0028] The vehicle control interface may further include a storage unit configured to store conversion information, the conversion information being a rule for converting a first control instruction into a second control instruction, wherein the control unit converts the first control instruction into the second control instruction based on the conversion information. For example, the storage unit pre-stores a rule (specific to the vehicle) for converting the first control instruction into the second control instruction, and generates a control instruction to be sent to the vehicle platform based on data sent from the autonomous driving platform. With such a configuration, the autonomous driving platform can be introduced regardless of the manufacturer or vehicle type.

[0029] The control unit may calculate a range of acceleration or deceleration or a range of steering angle variation that may be requested from the first computer based on information acquired from the vehicle platform.

[0030] The range of speed change (acceleration or deceleration) and steering angle change (angular velocity, etc.) per unit time that can be requested from the vehicle platform depends on the vehicle state and driving conditions (e.g., road conditions, traffic conditions, engine load conditions, number of occupants, etc.). Therefore, the vehicle control interface can calculate these based on information obtained from the vehicle platform. With this configuration, it is possible to determine whether the data sent from the autonomous driving platform is appropriate. In addition, the appropriate range can be notified to the autonomous driving platform.

[0031] If the acceleration or deceleration specified by the first data exceeds the range that can be requested from the first computer, the control unit may correct the acceleration or deceleration within a predetermined range. If the amount of change in the steering angle specified by the second data exceeds the range that can be requested from the first computer, the control unit may correct the amount of change in the steering angle within a predetermined range.

[0032] As described above, if the autonomous driving platform specifies inappropriate acceleration or deceleration or steering angle change, the vehicle control interface can automatically perform corrections. This configuration ensures safety.

[0033] The control unit can notify the second computer of the range of acceleration or deceleration or the range of steering angle change that can be requested from the first computer. As described above, the information required for making decisions during autonomous driving can be notified to the autonomous driving platform.

[0034] First embodiment

[0035] The outline of the vehicle system according to the first embodiment will be described. Figure 1 As shown, the vehicle system according to this embodiment is composed of a vehicle platform 100, an autonomous driving platform 200, and a vehicle control interface 300. Vehicle platform 100 is a conventional vehicle platform. It operates based on vehicle-specific control commands and generates vehicle-specific information. Control commands and vehicle information are encapsulated, for example, in Controller Area Network (CAN) frames that flow through an in-vehicle network.

[0036] The autonomous driving platform 200 has a device for sensing the surroundings of the vehicle and issues control instructions that are not specific to the vehicle type or manufacturer. In addition, vehicle information that is not specific to the vehicle type or manufacturer is obtained. The vehicle control interface 300 converts control instructions that are specific to the vehicle (i.e., control instructions that can be interpreted by the vehicle platform 100) and control instructions that are not specific to the vehicle (i.e., control instructions generated by the autonomous driving platform 200) into each other. In addition, the vehicle control interface 300 also converts vehicle information that is specific to the vehicle (i.e., vehicle information generated by the vehicle platform 100) and vehicle information that is not specific to the vehicle (i.e., vehicle information that can be interpreted by the autonomous driving platform 200) into each other.

[0037] Next, the components of the system will be described in detail. Figure 2 It is schematically shown Figure 1 FIG. 1 is a block diagram showing an example of a configuration of a vehicle system. The vehicle system includes a vehicle platform 100 , an autonomous driving platform 200 , and a vehicle control interface 300 , and each component is communicably connected via a bus 400 .

[0038] Vehicle platform 100 includes a vehicle control ECU 101, a braking system 102, a steering system 103, a steering angle sensor 111, and a vehicle speed sensor 112. This example uses a vehicle equipped with an engine, but an electric vehicle could also be used. In this case, the engine ECU could be replaced with an ECU that manages vehicle power. Furthermore, vehicle platform 100 could be equipped with ECUs and sensors different from those shown.

[0039] The vehicle control ECU 101 is a computer that controls various components of the vehicle (e.g., engine system components, powertrain components, brake system components, electrical system components, and body system components). The vehicle control ECU 101 may be a group of computers. The vehicle control ECU 101 controls the engine speed by, for example, executing fuel injection control. The vehicle control ECU 101 can control the engine speed based on control commands (e.g., commands for specifying the throttle opening) generated by, for example, an occupant's operation (e.g., operating the accelerator pedal).

[0040] If the vehicle is an electric vehicle, the vehicle control ECU 101 can control the motor speed by controlling the drive voltage, current, drive frequency, and other factors. In this case, as in the case of internal combustion engine vehicles, the motor speed can also be controlled based on control commands generated by occupant operation. Furthermore, the regenerative current can be controlled based on the force applied to the brake pedal and a control command indicating the degree of regenerative braking. If the vehicle is a hybrid vehicle, control can be performed for both the engine and the motor.

[0041] In addition, the vehicle control ECU 101 can control the braking force of the mechanical brake by controlling an actuator 1021 included in the brake device 102 described later. The vehicle control ECU 101 can control the brake hydraulic pressure by driving the actuator 1021 based on a control command (e.g., a command indicating a depression force on the brake pedal) generated by, for example, an operation of the occupant (e.g., operation of the brake pedal).

[0042] In addition, the vehicle control ECU 101 can control the steering angle or steering wheel angle by controlling a steering motor 1031 included in the steering device 103 described later. The vehicle control ECU 101 can control the steering angle of the vehicle by driving the steering motor 1031 based on a control command (e.g., a command indicating the steering angle) generated by, for example, an operation of the occupant (e.g., a steering operation).

[0043] The control command may be generated in the vehicle platform 100 based on the occupant's operation, or may be generated outside the vehicle platform 100 (eg, by a device that controls autonomous driving).

[0044] The brake device 102 is a mechanical brake system installed in the vehicle. It includes an interface (such as a brake pedal), an actuator 1021, a hydraulic system, a brake cylinder, and the like. The actuator 1021 is a device for controlling the hydraulic pressure in the brake system. The actuator 1021, having received instructions from the vehicle control ECU 101, controls the hydraulic pressure of the brake to ensure braking force of the mechanical brake.

[0045] The steering device 103 is a steering system provided in the vehicle. The steering device 103 includes interfaces such as a steering wheel, a steering motor 1031, a gearbox, and a steering column. The steering motor 1031 is a device for assisting the steering operation. The force required for the steering operation can be reduced by driving the steering motor 1031 after receiving a command from the vehicle control ECU 101. In addition, the steering operation can be automated by driving the steering motor 1031 without relying on the operation of the occupant.

[0046] The steering angle sensor 111 detects the steering angle obtained through steering operation. The detection value obtained by the steering angle sensor 111 is transmitted to the vehicle control ECU 101 as needed. In this embodiment, a numerical value directly indicating the tire rotation angle is used as the steering angle, but a value indirectly indicating the tire rotation angle may also be used. The vehicle speed sensor 112 detects the vehicle speed. The detection value obtained by the vehicle speed sensor 112 is transmitted to the vehicle control ECU 101 as needed.

[0047] The following describes the autonomous driving platform 200. The autonomous driving platform 200 senses the vehicle's surroundings, generates a driving plan based on the sensing results, and issues instructions to the vehicle platform 100 based on the plan. The autonomous driving platform 200 may be developed by a manufacturer or supplier different from that of the vehicle platform 100. The autonomous driving platform 200 includes an autonomous driving ECU 201 (as an example of a "first control unit") and a sensor group 202.

[0048] The autonomous driving ECU 201 is a computer that controls the vehicle by determining autonomous driving based on data acquired from the sensor group 202, described later, and by communicating with the vehicle platform 100. The autonomous driving ECU 201 is composed of, for example, a CPU (Central Processing Unit). The autonomous driving ECU 201 includes two functional modules: a situation recognition unit 2011 and an autonomous driving control unit 2012. Each functional module can be implemented by the CPU executing a program stored in a storage unit such as a ROM (Read Only Memory).

[0049] The situation recognition unit 2011 detects the environment surrounding the vehicle based on data acquired by sensors included in the sensor group 202, described later. Detection targets include, but are not limited to, the number and location of lanes, the number and location of vehicles around the vehicle, the number and location of obstacles (e.g., pedestrians, bicycles, structures, buildings, etc.) around the vehicle, road structure, road signs, and the like. Any detection target can be used as long as it is necessary for autonomous driving. Data related to the environment detected by the situation recognition unit 2011 (hereinafter referred to as "environmental data") is transmitted to the autonomous driving control unit 2012.

[0050] The autonomous driving control unit 2012 uses the environmental data generated by the situation recognition unit 2011 to control the vehicle's travel. For example, the control unit generates a driving trajectory for the vehicle based on the environmental data and determines the vehicle's acceleration or deceleration and steering angle to ensure the vehicle follows the driving trajectory. The information determined by the autonomous driving control unit 2012 is transmitted to the vehicle platform 100 (vehicle control ECU 101) via the vehicle control interface 300, described later. Known methods can be employed to enable the vehicle to travel autonomously.

[0051] In this embodiment, the automatic driving control unit 2012 only generates instructions related to the acceleration or deceleration of the vehicle and instructions related to the steering of the vehicle as first control instructions. Hereinafter, instructions related to the acceleration or deceleration of the vehicle are referred to as acceleration and deceleration instructions, while instructions related to the steering of the vehicle are referred to as steering instructions. The first control instructions, which include acceleration and deceleration instructions and steering instructions, are general instructions that are not dependent on the type or manufacturer of the vehicle. In this embodiment, the acceleration and deceleration instructions are information that specifies the acceleration or deceleration of the vehicle, while the steering instruction is information that specifies the steering angle of the vehicle's steering wheel.

[0052] The sensor group 202 is a unit configured to sense the vehicle's surroundings and typically includes a monocular camera, a stereo camera, a radar, a laser radar (LIDAR), a laser scanner, and the like. In addition to these devices for sensing the vehicle's surroundings, the sensor group 202 may also include a device for acquiring the vehicle's current location (e.g., a GPS module). Information acquired by the sensors included in the sensor group 202 is transmitted to the autonomous driving ECU 201 (situation recognition unit 2011) as needed.

[0053] Next, the vehicle control interface 300 will be described. In this embodiment, the control instructions processed by the vehicle control ECU 101 are specific to the vehicle and manufacturer. On the other hand, the autonomous driving platform 200 is a device developed by a third party and is expected to be installed in a variety of vehicle types from various manufacturers. In other words, connecting both components to the same in-vehicle network is expensive. Therefore, in this embodiment, the vehicle control interface 300 serves as a device for converting and relaying data exchanged between the vehicle control ECU 101 and the autonomous driving ECU 201.

[0054] The control unit 301 (as an example of a "third control unit") is a computer that converts the control instructions processed by the vehicle control ECU 101 and the control instructions processed by the automatic driving ECU 201. The control unit 301 is composed of a CPU (Central Processing Unit), for example. Figure 3 As shown, the control unit 301 includes three functional modules, an acceleration / deceleration instruction processing unit 3011, a steering instruction processing unit 3012, and a vehicle information processing unit 3013. Each functional module can be implemented by the CPU executing a program stored in the storage unit 302 described later.

[0055] The acceleration / deceleration command processing unit 3011 receives acceleration / deceleration commands from the autonomous driving ECU 201 and converts them into data (second control commands; hereinafter referred to as "control data") that can be interpreted by the vehicle control ECU 101. Specifically, the acceleration or deceleration specified by the acceleration / deceleration command (e.g., +3.0 km / h / s) is converted into data indicating the throttle opening or data indicating the brake pressure. The converted control data is transmitted using a protocol or format specific to the vehicle platform 100. The conversion process is performed using conversion information stored in the storage unit 302, which will be described later. This process will be described later.

[0056] In this example, the throttle opening and the brake pressure are exemplified as the control data. However, the control data may be other data as long as it is related to the acceleration or deceleration of the vehicle. For example, the target speed or current value of the electric motor may be used.

[0057] The steering command processing unit 3012 receives steering commands from the autonomous driving ECU 201 and, using conversion information, converts the steering commands into control data that can be interpreted by the vehicle control ECU 101. Specifically, the data is converted into data indicating the steering angle specific to the vehicle platform 100. In this example, the tire angle is illustrated as the steering angle, but the control data may be other data as long as it is relevant to vehicle steering. For example, the control data may directly or indirectly represent the steering wheel angle, a percentage of the maximum steering angle, or the like.

[0058] The vehicle information processing unit 3013 receives information about the vehicle state from the vehicle control ECU 101 and converts the information into information that can be interpreted by the autonomous driving ECU 201 (information that is not specific to the vehicle type). In particular, information sent in a protocol or format specific to the vehicle platform 100 is converted into information in a universal format (hereinafter referred to as feedback data). In the following, information about the vehicle state is referred to as sensor data. The sensor data is based on information obtained by the steering angle sensor 111 and the vehicle speed sensor 112, for example, and is sent by the vehicle control ECU 101 to the on-board network. For example, the sensor data can be any data as long as it can be fed back to the autonomous driving ECU 201, such as vehicle speed information, information about the turning angle of the tires, and information about the steering angle. In this embodiment, the vehicle information processing unit 3013 converts sensor data related to the current vehicle speed and steering angle state.

[0059] The storage unit 302 is a unit configured to store information and is composed of a storage medium such as RAM, a magnetic disk, or a flash memory. The storage unit 302 stores information (hereinafter referred to as conversion information) used to convert acceleration and deceleration commands and steering commands generated by the autonomous driving ECU 201 (autonomous driving control unit 2012) into control data interpretable by the vehicle control ECU 101 (and vice versa). This conversion information also includes information used to convert vehicle-specific sensor data into feedback data.

[0060] The conversion information includes, for example, the configuration of control data input to or output from the vehicle control ECU 101, its parameters, and a table or mathematical formula for converting the input values ​​into the parameters. Furthermore, the conversion information is composed of the configuration of sensor data output from the vehicle control ECU 101, its parameters, a table or mathematical formula for converting the parameters into physical values, and the like.

[0061] Figure 4 1 is a diagram showing the types of data converted by conversion information. In the diagram, "input" indicates that it is data from the autonomous driving ECU 201 to the vehicle control ECU 101, and "output" indicates that it is data from the vehicle control ECU 101 to the autonomous driving ECU 201. As described above, instructions related to acceleration or deceleration and steering angle are sent from the autonomous driving ECU 201 to the vehicle control ECU 101, and data related to the current vehicle speed and steering angle state are sent from the vehicle control ECU 101 to the autonomous driving ECU 201. Figure 4 When data other than the data shown is transmitted to the vehicle control interface 300 , the data is discarded.

[0062] In the vehicle system according to the present embodiment, communication between the vehicle platform 100 and the autonomous driving platform 200 is performed by the above-described configuration.

[0063] Next, we will refer to Figure 5A and Figure 6A The processing performed by the vehicle system according to the present embodiment is described with reference to a processing flowchart of FIG. Figure 5A The illustrated process is executed by the autonomous driving platform 200 at predetermined intervals.

[0064] In step S11, the automatic driving ECU 201 generates a driving plan based on the information acquired from the sensor group 202. The driving plan is data indicating the behavior of the vehicle within a predetermined interval. Figure 7 As shown, when a driving plan is generated for a vehicle traveling in a first lane to move to a second lane, a driving trajectory is generated as shown in the figure. The driving plan may include the vehicle's driving trajectory or information related to vehicle acceleration or deceleration. A driving plan may also be generated based on information other than the exemplary information. For example, a driving plan may be generated based on a departure point, via points, a destination, map data, and the like.

[0065] In step S12, the automatic driving ECU 201 generates physical control variables for implementing the driving plan. In this embodiment, two types of physical control variables are generated: a physical control variable for acceleration or deceleration and a physical control variable for the steering angle. Figure 8A is a timing chart showing the control amount for acceleration or deceleration, Figure 8B This is a timing diagram showing the control amount for the steering angle. Each value can be generated based on pre-set parameters, such as the relationship between vehicle speed and maximum steering angle, the relationship between driving conditions and acceleration or deceleration (steering angle), or the time period required to complete a maneuver (e.g., changing lanes).

[0066] In step S13, the automatic driving ECU 201 divides each generated physical control amount into a plurality of time steps. The time step may be, for example, 100 milliseconds, but is not limited thereto. Figure 8C An example is shown in which the physical control amount for the generated acceleration or deceleration is divided into seven steps within the period from time t1 to time t2.

[0067] In step S14, the automatic driving ECU 201 calculates the current time step t based on the physical control amount. n To the next time step t n+1Acceleration / deceleration commands and steering commands are issued based on changes in the steering angle. For example, when a time step is 100 milliseconds and +2.0 km / h / s is specified as the acceleration or deceleration, an acceleration / deceleration command is generated that specifies a change of 0.2 km / h per time step. For example, when a steering angle change of 20 degrees within 2 seconds is specified, a steering command is generated that specifies a change of 1 degree per time step. The generated acceleration / deceleration commands and steering command are input to the control unit 301 of the vehicle control interface 300.

[0068] In step S15, the vehicle control interface 300 (control unit 301) processes the acquired acceleration / deceleration commands and steering commands. Figure 6 is a diagram illustrating the processing in step S15 in detail. In step S151, the acceleration / deceleration command processing unit 3011 acquires the acceleration / deceleration commands transmitted from the autonomous driving ECU 201. Similarly, in step S152, the steering command processing unit 3012 acquires the steering command transmitted from the autonomous driving ECU 201.

[0069] In step S153, control unit 301 performs data conversion. Specifically, acceleration / deceleration command processing unit 3011 performs conversion between acceleration / deceleration commands and control data based on the conversion information stored in storage unit 302. The control data to be converted is data indicating the throttle opening or data specifying the brake pressure. Furthermore, steering command processing unit 3012 performs conversion between steering commands and control data based on the conversion information stored in storage unit 302. The control data to be converted is data indicating the steering angle (the turning angle of the tires).

[0070] In step S154, the generated control data is sent to the vehicle control ECU 101. In this step, for example, the control data generated in step S153 is encapsulated in a data frame transmitted or received by the vehicle network and is sent to the vehicle control ECU 101 as the destination. In addition, in step S15, the vehicle control interface 300 receives the data except Figure 4 If data other than the indicated data is present, the data is discarded.

[0071] Will return Figure 5A Continuing the description, step S16 is a step in which the autonomous driving ECU 201 senses the vehicle state after transmitting the control data. In this step, the sensor data transmitted from the vehicle control ECU 101 is converted by the vehicle control interface 300 based on the conversion information and then relayed to the autonomous driving ECU 201. Upon receiving this data, the autonomous driving ECU 201 determines whether the vehicle is in the desired state.

[0072] Because vehicle behavior is affected by factors such as the current engine load and road conditions (e.g., slope), in this embodiment, the autonomous driving ECU 201 receives feedback from sensor data and determines whether the required physical control variables have been achieved. The sensor data is acquired by the vehicle information processing unit 3013, converted into feedback data (indicating the current vehicle speed and steering angle), and then sent to the autonomous driving ECU 201. Figure 3 and Figure 4 While exemplary data indicating the current vehicle speed and steering angle are shown as feedback data, the feedback data is not limited thereto. For example, the feedback data may include data related to factors influencing vehicle behavior, such as tire angle, steering angle, angular velocity, engine load, road grade (inclination), number of passengers, load capacity, road conditions, and traffic conditions.

[0073] In step S17, the autonomous driving ECU 201 corrects the driving plan based on the received feedback data. For example, if the feedback data indicates that the engine load is high and the requested acceleration cannot be achieved, the driving plan is corrected to achieve higher acceleration. Furthermore, while this example illustrates correcting the driving plan, there may be cases where the driving control cannot be changed, but the physical control variables used to implement the driving plan can be corrected.

[0074] In the vehicle system according to the first embodiment, by executing the above-described processing, appropriate vehicle driving control can be performed according to the vehicle's condition. Specifically, by defining the data to be relayed by vehicle control interface 300 as instructions related to acceleration and deceleration and instructions related to steering, and filtering other instructions, access to unnecessary vehicle functions can be prevented and safety ensured. Furthermore, by preparing conversion information, autonomous driving platform 200 can be applied to various vehicle types without modification.

[0075] In the description of this embodiment, the autonomous driving ECU 201 corrects the difference between the actual state of the vehicle and the ideal state of the vehicle based on the feedback data. However, the vehicle control interface 300 can also perform the correction. For example, the feedback data generated by the vehicle information processing unit 3013 can be input to the acceleration and deceleration instruction processing unit 3011 (steering instruction processing unit 3012), so that the acceleration and deceleration instruction processing unit 3011 (steering instruction processing unit 3012) automatically corrects the control data. In addition, the autonomous driving ECU 201 can generate data specifying the amount to be corrected independently of the acceleration and deceleration instructions and the steering instructions, and can send the data to the vehicle control interface 300.

[0076] In the description of this embodiment, the autonomous driving ECU 201 sends two types of instructions (i.e., acceleration and deceleration instructions and steering instructions) to the vehicle control interface 300, but other information can be sent as additional information. In addition, the vehicle control interface 300 can generate control data to be sent to the vehicle control ECU 101 based on the additional information. In the description of the embodiment, the steering angle (the turning angle of the tires) is used as the steering instruction. However, the steering instruction can also be information about the trajectory of the vehicle itself.

[0077] Second embodiment

[0078] The following will refer to Figure 5B and Figure 6B The processing performed by the vehicle system according to the present embodiment is described with reference to a processing flowchart of FIG. Figure 5B The process shown may be performed by the autonomous driving platform 200 at predetermined intervals. Figure 5B As shown, at step S161, it can be determined that the sensor (for example but not limited to Figure 2 At least one of the following information, including the movement of an object detected by the sensor group 202 (e.g., a camera, radar, lidar, etc.), its positional relationship with a reference object, and its distribution, is used as obstacle-related information. The obstacle-related information can then be used to determine whether the detected object is an obstacle (step S162). If the detected object is determined not to be an obstacle (the determination result of step S162 is negative), the vehicle's driving is not changed (step S171). For example, step S171 may generate a first control instruction that does not correct the driving plan. If the detected object is determined to be an obstacle (i.e., the determination result of step S162 is positive), the process proceeds to step S17 to correct the driving plan.

[0079] Return Reference Figure 2The above-described processing will be specifically described with reference to the configuration of the autonomous driving platform 200 shown in FIG. The autonomous driving ECU 201, as an example of a first control unit, can be configured to execute each step of the above-described processing (and examples of processing in various embodiments of the present disclosure). Furthermore, the sensor group 202 can be configured to collect environmental data surrounding the vehicle (e.g., information about objects surrounding the vehicle). For example, radar or lidar can be used to collect reflection intensity data of various objects surrounding the vehicle, or a camera can be used to capture static or dynamic images (video) of the vehicle's surroundings. The situation recognition unit 2011 can be configured to: determine at least one of the motion status of a detected object, its positional relationship with a reference object, and its distribution status as obstacle-related information based on the environmental data acquired by the sensors included in the sensor group 202; and determine whether the object is an obstacle based on the obstacle-related information. The information determined by the situation recognition unit 2011 as an obstacle can be transmitted to the autonomous driving control unit 2012. The autonomous driving control unit 2012 thus uses the environmental data generated by the situation recognition unit 2011 to control the vehicle's travel. Specifically, upon receiving information from the situation recognition unit 2011 that surrounding objects are not obstacles, the autonomous driving control unit 2012 may generate a first control instruction that does not alter the vehicle's driving direction. The above is merely an example; some of the functions implemented by the situation recognition unit 2011 may also be implemented by the autonomous driving control unit 2012. In this way, by considering at least one of the motion state of objects easily misidentified as obstacles (hereinafter referred to as "pseudo-" obstacles), their positional relationship with reference objects, and their distribution relative to obstacles, "pseudo-" obstacles can be accurately and quickly distinguished from obstacles, significantly reducing the error rate of obstacle detection. This effectively prevents the vehicle from undergoing unnecessary changes in driving direction due to misidentified obstacles, thereby improving the autonomous driving experience.

[0080] In some embodiments, the first control instruction that does not correct the driving plan can be converted into a second control instruction that can be interpreted by the vehicle platform 100 via the vehicle control interface 300, and sent to the vehicle control ECU 101 of the vehicle platform, so that it can control and drive the braking device 102, steering device 103, etc. to continue to execute the previous driving plan. Figure 6B The vehicle control interface 300 can operate as follows. In step S155, a first control instruction that does not change the vehicle's driving behavior is obtained. In step S156, the first control instruction is converted into a second control instruction specific to the vehicle platform 100. In step S157, a data frame specific to the ECU of the vehicle platform 100 is generated. For example, the control data of the second control instruction is encapsulated in a data frame specific to the ECU of the vehicle platform 100, and the data frame is transmitted to the ECU of the vehicle platform 100 via, for example, an in-vehicle network.

[0081] In some embodiments, Figure 5B The processing shown in Figure 5A , steps S16 and S17 are integrated. The automatic driving method of the vehicle according to the present disclosure can achieve good identification effects for various common "pseudo" obstacles. Various forms of obstacle-related information can be adopted according to different needs. In some embodiments, the motion condition of the object may include the floating condition of the object relative to the detection space and the positional relationship between the object and the reference object may include the positional relationship between the object and the ground, so that the two together are used as obstacle-related information. In some embodiments, the motion condition of the object may include the motion vector of the object relative to the detection space, so that it is used as obstacle-related information. In some embodiments, the positional relationship between the object and the reference object may include the positional relationship between the object and the lane and the distribution condition of the object may include the continuous distribution condition of the reflection intensity of the object, so that the two together are used as obstacle-related information. For various application scenarios, the motion condition of the object, the positional relationship with the reference object, and the distribution condition of the object in a suitable form can be adopted as obstacle-related information, or a suitable combination of these factors can be selected.

[0082] The following details the processing of this automated driving method using "floating objects," "floating animals," and fast-growing roadside plants as examples of "pseudo" obstacles. However, it should be noted that these are merely examples. The object's motion, its positional relationship to reference objects, and its distribution can provide reference information for effectively detecting various "pseudo" obstacles in various application scenarios.

[0083] As used herein, the technical term "floating objects" refers to objects floating in the air, such as dust, dirt, and particulate matter (pollen). With the widespread application of new high-resolution sensors such as lidar in vehicle platform 100 or autonomous driving platform 200, these sensors can sometimes detect tiny "floating objects" that are invisible to conventional sensors. This can cause the vehicle platform 100 or autonomous driving platform 200 to mistake these objects for obstacles and alter its driving strategy, such as slowing down or maneuvering. This can severely impact the autonomous driving experience and even cause safety issues.

[0084] In some embodiments, the object's motion status can include its floating state relative to the detection space, and the positional relationship between the object and a reference object can include its positional relationship with the ground, and together these can serve as obstacle-related information. For example, in an application scenario for detecting floating objects, it can be determined whether the object's size is less than a first threshold (which can be preset based on the size of the floating object) and whether it is not in contact with the ground. If so, the object can be determined to be a tiny particle and airborne. Furthermore, a detection space adapted to the size of the floating object can be preset, and the object's retention relative to the detection space over a period of time can be determined. The inventors discovered that airborne obstacles can also appear to a lidar as tiny and airborne, such as highly reflective points on a dark object. These points remain largely stationary relative to the detection space, whereas floating objects in the air are easily disturbed and do not remain in the detection space. By further combining the object's retention relative to the detection space over a period of time, floating objects can be distinguished from obstacles and accurately detected, thereby avoiding misidentification of various floating objects as obstacles.

[0085] Herein, the technical term "floating object" is intended to refer to objects that are relatively larger than "floating objects" and exhibit fluttering motion, such as balloons, plastic bags, flags, catkins, and the like. Taking the application scenario of detecting floating objects as an example, the object's motion status can also include its motion vector relative to the detection space, thereby serving as obstacle-related information. Note that the scale of the detection space can be adjusted to suit the object's size, so that the object's motion status relative to the detection space accurately reflects the object's inherent motion characteristics. For example, the size of the detection space for dust can be set to reflect the dust's continuous wandering motion characteristics, while the size of the detection space for a plastic bag can be set to reflect the plastic bag's continuous, irregular movement in the air. Specifically, if the object's motion vector remains away from the ground (such as a balloon flying away) or changes direction multiple times over a period of time (such as a plastic bag drifting erratically in the wind), the object can be determined to be a floating object rather than an obstacle. This effectively prevents various floating objects from being misidentified as obstacles.

[0086] In some application scenarios, the autonomous driving method will refer to the on-board map to plan the driving trajectory, but the on-board map may not record real-time and sufficiently detailed road conditions. For example, flowers and crops are often planted near the edge of the lane. These flowers and crops grow rapidly, and therefore are usually not marked and updated in a timely manner in the on-board map. The sensor group 202 of the autonomous driving platform 200, such as radar or lidar, will detect the reflection intensity related data of these flowers and crops. In some embodiments, the positional relationship between the object and the reference object may include the positional relationship between the object and the lane, and the distribution of the object may include the continuous distribution of the reflection intensity of the object, which can be used together as obstacle-related information. Specifically, in the application scenario of detecting flowers and plants on the roadside, it can be determined whether the object is adjacent to the edge of the lane and whether it is continuously distributed with a reflection intensity lower than the second threshold. If so, it can be determined that the object is a flower and plant rather than an obstacle. In this way, it is possible to effectively avoid misjudging these flowers and plants as obstacles.

[0087] The various examples of processing above are merely illustrative of the autonomous driving method disclosed herein and may be performed by, for example, Figure 2 This is achieved by the autonomous driving platform 200 shown in FIG, which will not be described in detail here.

[0088] Third embodiment

[0089] In the first embodiment, the vehicle control interface 300 performs data conversion based on the conversion information stored in the storage unit 302. However, depending on the vehicle state, this conversion may be inappropriate without changing the command sent from the autonomous driving platform 200. The third embodiment is an embodiment that limits the range of acceleration or deceleration and steering angle to address this problem.

[0090] The configuration of the vehicle system according to the third embodiment is the same as that of the first embodiment, except that the vehicle control interface 300 (vehicle information processing unit 3013) has the function of generating information about the range of acceleration or deceleration and the range of steering angle (hereinafter referred to as range information) that can be specified based on sensor data obtained from the vehicle platform and notifying the range information to the autonomous driving platform.

[0091] In the third embodiment, the vehicle information processing unit 3013 calculates the range of acceleration or deceleration that can be specified, as well as the range of steering angles, based on acquired sensor data, and notifies the autonomous driving ECU 201 of these ranges. For example, the achievable vehicle acceleration or deceleration can be varied depending on the number of occupants, load capacity, engine load, road conditions, and the like. Furthermore, the achievable steering angle range can be varied depending on vehicle speed, road conditions, traffic conditions, and the like. By calculating these ranges and notifying the autonomous driving ECU 201 of the range data, appropriate control can be implemented.

[0092] Examples of the range information to be notified include the following: (1) a range of acceleration or deceleration that can be specified (lower and upper limits); (2) a range of steering angles that can be specified (left and right angles); (3) a range of steering angle changes (angular velocities) that can be specified; and (4) a range of lateral acceleration or lateral jerk. This information is estimated and generated based on sensor data. Rules for generating range information are pre-stored in storage unit 302.

[0093] The range information generated by the vehicle control interface 300 is transmitted to the autonomous driving ECU 201 and used in steps S11 and S12. For example, in step S12, the physical control variables are generated so that the acceleration or deceleration, steering angle, and angular velocity of the steering angle fall within the notified range. Alternatively, in step S11, a driving plan is generated so that the physical control variables do not fall outside of the range.

[0094] Furthermore, in the third embodiment, if the acceleration / deceleration commands and steering commands generated by the autonomous driving ECU 201 exceed the aforementioned ranges, those commands are corrected. For example, if the acceleration / deceleration commands and steering commands generated by the autonomous driving ECU 201 include values ​​exceeding an upper limit (lower limit), control data is generated assuming that the upper limit (lower limit) has been specified. Consequently, the vehicle can be controlled only within a range deemed safe by the vehicle.

[0095] In addition, in the case where corrections are made based on the range information, the feedback sent to the autonomous driving ECU 201 may include that the corrections have been made. Therefore, the autonomous driving ECU 201 can regenerate the driving plan.

[0096] In the description of this embodiment, the vehicle information processing unit 3013 generates information about vehicle speed and steering angle as range data, but other information may be added. For example, an estimated value of acceleration or deceleration when the throttle is fully closed may be added to the range data.

[0097] Modify the example

[0098] The above-described embodiments are merely examples, and the present invention can be implemented with appropriate modifications within the scope of the gist thereof. For example, unless a technical contradiction occurs, the processes and units described in the present disclosure can be freely combined and implemented.

[0099] Furthermore, a process described as being performed by a single device may be performed by multiple devices in a shared manner. Alternatively, a process described as being performed by different devices may be performed by a single device. In a computer system, the hardware configuration (server configuration) used to implement each function can be flexibly changed.

[0100] The present invention can also be implemented by providing a computer program for executing the functions described in the embodiments in a computer, and reading and executing the program by one or more processors included in the computer. Such a computer program can be provided to the computer via a non-transitory computer-readable storage medium that can be connected to the computer system bus, or can be provided to the computer via a network. Examples of non-transitory computer-readable storage media include random access disks (e.g., disks ( disks, hard disk drives (HDDs) and optical disks (CD-ROMs, DVD disks, Blu-ray discs, etc.), read-only memories (ROMs), random-access memories (RAMs), EPROMs, EEPROMs, magnetic cards, flash memories, optical cards, and random types of media suitable for storing electronic instructions.

Claims

1. A method for automatic driving of a vehicle, characterized in that: The autonomous driving method comprises: determining at least one of a motion status of an object detected by a sensor, a positional relationship with a reference object, and a distribution status of the object as obstacle-related information; determining whether the object is an obstacle based on the obstacle-related information; and If it is determined that the object is not an obstacle, the driving of the vehicle is not changed; The obstacle-related information includes a motion vector of the object relative to the detection space; Determining whether the object is an obstacle based on the obstacle-related information includes: determining that the object is a floating object rather than an obstacle when the motion vector remains away from the ground or changes direction multiple times within a period of time.

2. The automatic driving method according to claim 1, wherein: The obstacle-related information includes the floating status of the object relative to the detection space and the positional relationship between the object and the ground; Determining whether the object is an obstacle based on the obstacle-related information includes: determining that the object is a floating object rather than an obstacle if the size of the object is smaller than a first threshold, is not in contact with the ground, and does not float in the detection space for a period of time.

3. The automatic driving method according to claim 1 or 2, characterized in that: Also includes: The scale of the detection space is adjusted according to the size of the object.

4. The automatic driving method according to claim 1, wherein: The obstacle-related information includes the positional relationship between the object and the lane and the continuous distribution of the reflection intensity of the object; Determining whether the object is an obstacle based on the obstacle-related information includes: determining that the object is a plant rather than an obstacle when the object is adjacent to an edge of a lane and is continuously distributed with a reflection intensity lower than a second threshold.

5. An autonomous driving platform comprising: a sensor group configured to detect objects around the vehicle; as well as a first control unit for performing automatic driving control of the vehicle, configured to: determining, based on the information detected by the sensor group, at least one of a motion status of the detected object, a positional relationship with a reference object, and a distribution status of the object as obstacle-related information; determining whether the object is an obstacle based on the obstacle-related information; and When it is determined that the object is not an obstacle, issuing a first control instruction that does not change the driving of the vehicle; The first control unit is further configured to determine that the object is a floating object rather than an obstacle when the motion vector of the object relative to the detection space remains away from the ground or changes direction multiple times within a period of time.

6. The autonomous driving platform according to claim 5, characterized in that: The sensor group includes a laser radar, and the first control unit is further configured to: In a case where the size of the object is smaller than a first threshold, the object is not in contact with the ground, and the object does not float in the detection space for a period of time, it is determined that the object is a floating object rather than an obstacle.

7. The autonomous driving platform according to claim 5, characterized in that: The sensor group includes at least one of a radar and a lidar, and the first control unit is further configured to determine that the object is a flower or grass rather than an obstacle when the object is adjacent to the edge of the lane and is continuously distributed with a reflection intensity lower than a second threshold.

8. A vehicle system for use in conjunction with a vehicle platform, the vehicle platform including a second control unit that performs driving control of the vehicle, characterized in that: The vehicle system includes: The autonomous driving platform according to any one of claims 5 to 7; and a vehicle control interface configured to connect the vehicle platform and the autonomous driving platform, wherein the vehicle control interface includes a third control unit configured to perform: acquiring the first control instruction from the first control unit; converting the first control instruction into a second control instruction for the second control unit; and The second control instruction is sent to the second control unit.

Images