An obstacle avoidance path generation method for an intelligent networked low-speed teaching vehicle

Abstract

This invention relates to the field of intelligent connected low-speed training vehicles, and discloses a method for generating obstacle avoidance paths for such vehicles. The method includes: acquiring vehicle positioning information, a predetermined driving path, and preset obstacle avoidance decision parameters, including obstacle avoidance distance, safety factor, obstacle avoidance path threshold, detour distance threshold, and filtering coefficient; when a static obstacle is detected on the predetermined driving path, acquiring its position and size information, and converting the position information to the vehicle coordinate system; determining the distance information between the vehicle and the static obstacle, determining the obstacle avoidance starting position based on the obstacle avoidance distance and distance information, and calculating the deflection angle based on the size information, distance information, and safety factor; generating an obstacle avoidance path based on the obstacle avoidance starting position, deflection angle, obstacle avoidance path threshold, detour distance threshold, and filtering coefficient, and generating a return path after the vehicle bypasses the static obstacle. This achieves automatic obstacle avoidance of static obstacles for intelligent connected low-speed training vehicles.

Description

Technical Field

[0001] This invention relates to the field of intelligent connected low-speed teaching vehicles, and specifically to a method for generating obstacle avoidance paths for intelligent connected low-speed teaching vehicles. Background Technology

[0002] With the development of intelligent technologies, intelligent connected vehicles have become an important development direction for the automotive industry. Intelligent connected vehicles integrate internet technology, automatic control technology, environmental perception technology, and big data technology to achieve vehicle acquisition, processing, and control decisions based on driving environment information. For example, vehicles can acquire surrounding environmental information through sensing devices such as LiDAR, millimeter-wave radar, and cameras. This data is then processed by the positioning system and onboard processor, and combined with the vehicle control system to achieve autonomous driving functions. In an autonomous driving system, when a vehicle is traveling on a predetermined path, if an obstacle appears ahead, a path planning algorithm is needed to generate a new trajectory so that the vehicle can bypass the obstacle and continue driving. In existing technologies, autonomous vehicles typically rely on high-precision maps or SLAM (Simultaneous Localization and Mapping) technology to construct an environmental map, and combine this with path planning methods such as A*Sat algorithm and Bézier curves to generate the vehicle's driving trajectory. On the other hand, in the teaching and practical training of intelligent connected vehicle-related majors, especially in the whole-vehicle training of secondary vocational schools, higher vocational colleges, and applied undergraduate colleges, the teaching vehicles are usually low-speed modified vehicles, mainly used for experimental demonstrations and teaching practice of autonomous driving-related technologies. For example, in the autonomous driving teaching process, the teaching vehicles usually perform autonomous driving tasks based on GPS maps or predetermined driving routes to demonstrate the operation of the autonomous driving system in a real vehicle to students. However, in the current teaching practice, secondary vocational schools, higher vocational colleges, and applied undergraduate colleges generally lack practical training content related to vehicle obstacle avoidance in the whole-vehicle testing teaching of autonomous driving, such as the teaching content on obstacle avoidance decision-making process, obstacle avoidance path planning, and vehicle obstacle avoidance test case design. It is difficult to effectively combine test case design with whole-vehicle decision-making and path planning, thus making it difficult to form a complete autonomous driving practical training teaching system. At the same time, since the teaching vehicles are mostly modified vehicles, including multiple subsystems such as brake-by-wire system, drive-by-wire system, and steering-by-wire system, when anomalies occur during autonomous driving, it is not easy to distinguish whether the problem originates from the test case design, the decision-making process, or the vehicle hardware system, thus affecting the relevant teaching and training effects.

[0003] Therefore, in the teaching and training of autonomous driving, there is a lack of corresponding obstacle avoidance path planning teaching schemes when vehicles encounter static obstacles in front of them during driving. It is difficult to intuitively demonstrate the obstacle avoidance decision-making and path planning process in autonomous driving, thus affecting the teaching and experimental effects of autonomous driving-related technologies. Summary of the Invention

[0004] The summary section of this invention provides a brief overview of the concepts, which will be described in detail in the detailed description section below. This summary section is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0005] This invention proposes a method for generating obstacle avoidance paths for intelligent connected low-speed teaching vehicles to solve one or more of the technical problems mentioned in the background section above.

[0006] This invention provides a method for generating obstacle avoidance paths for an intelligent connected low-speed teaching vehicle, including: acquiring the vehicle's positioning information, predetermined driving path information, and preset obstacle avoidance decision parameters, wherein the preset obstacle avoidance decision parameters include obstacle avoidance distance, safety factor, obstacle avoidance path threshold, detour distance threshold, and filtering factor. Detect whether there are static obstacles on the predetermined driving path; when a static obstacle is detected on the predetermined driving path, obtain the position and size information of the static obstacle, and convert the position information of the static obstacle to the vehicle coordinate system; Based on the position information of the static obstacle in the vehicle coordinate system, determine the distance information between the vehicle and the static obstacle; Determine the starting position for obstacle avoidance based on the obstacle avoidance distance and distance information; Based on the size information, distance information, and safety factor of the static obstacle, calculate the deflection angle of the obstacle avoidance path relative to the predetermined driving path; Based on the obstacle avoidance starting position, deflection angle, obstacle avoidance path threshold, detour distance threshold, and filtering coefficient, an obstacle avoidance path is generated to bypass static obstacles. After the vehicle bypasses a static obstacle, a regression path is generated to return to the predetermined driving path.

[0007] Optionally, the obstacle avoidance starting position can be determined based on the obstacle avoidance distance and distance information, including: When the distance information is less than the obstacle avoidance distance, the obstacle avoidance condition is triggered, and the current vehicle position is used as the starting tracking point of the obstacle avoidance path.

[0008] Optionally, generating obstacle avoidance paths around static obstacles includes: The end point of the obstacle avoidance path is determined based on the obstacle avoidance path threshold.

[0009] Optionally, the size information of static obstacles includes the obstacle length and obstacle width.

[0010] Optionally, generating obstacle avoidance paths around static obstacles also includes: Calculate the straight-line detour distance based on the obstacle length and the detour distance threshold; Starting from the end of the tracking point, generate a detour segment with a length equal to the straight-line distance of the detour and parallel to the predetermined driving path.

[0011] Optionally, generating obstacle avoidance paths around static obstacles also includes: Based on the filtering coefficient and the preset default distance, the longitudinal step distance between tracking points is determined, where the preset default distance is the distance between adjacent tracking points in the given driving path.

[0012] Optionally, generating obstacle avoidance paths around static obstacles also includes: Between the starting position and the ending tracking point of obstacle avoidance, the longitudinal coordinate is maintained by a longitudinal step distance, and the change in the lateral coordinate is determined based on the longitudinal step distance and the deflection angle to generate the obstacle avoidance segment.

[0013] Optionally, the change in the lateral coordinate is determined based on the longitudinal step distance and the deflection angle, including: The lateral coordinate increment is determined based on the sine value of the deflection angle and the longitudinal step distance; Based on the vehicle's obstacle avoidance direction, the current lateral coordinate is offset using the lateral coordinate increment.

[0014] Optionally, a regression path is generated to return to the predetermined driving path, including: Using the end point of the obstacle avoidance path as the starting point, the longitudinal coordinate is maintained at a preset step distance, and the regression change of the lateral coordinate is determined according to the preset step distance and deflection angle to generate a regression segment. The lateral coordinate change direction of the regression segment is opposite to that of the obstacle avoidance segment.

[0015] This invention offers the following advantages: Addressing the need for path planning and obstacle avoidance training in autonomous driving teaching scenarios using intelligent connected low-speed training vehicles, when static obstacles appear on the vehicle's predetermined driving path, obstacle avoidance-related decision parameters are set. This allows the vehicle to dynamically generate an obstacle avoidance path based on the obstacle's size and the distance between the vehicle and the obstacle, enabling the vehicle to bypass the obstacle and ultimately return to the predetermined driving path in an autonomous driving process. Furthermore, the generation of the obstacle avoidance path does not rely on SLAM maps or high-precision maps, nor on complex path planning algorithms. Instead, by setting decision parameters such as a safety factor, detour distance threshold, filtering factor, obstacle avoidance distance, and obstacle avoidance path threshold, the obstacle avoidance path is dynamically generated based on the positional relationship between the vehicle and the static obstacle. This allows the vehicle to achieve obstacle avoidance functionality against static obstacles under relatively simple system conditions. Meanwhile, by adjusting the above decision parameters, various obstacle avoidance paths can be generated, enabling instructors to design different autonomous driving test cases based on vehicle speed, obstacle size, and testing requirements. By comparing and analyzing the actual obstacle avoidance results of the vehicle under different parameter settings, the perception information processing and decision-making process of the autonomous driving system during obstacle avoidance can be more intuitively demonstrated, which is conducive to the teaching and experimentation of autonomous driving-related technologies. Attached Figure Description

[0016] The above and other features, advantages, and aspects of the various embodiments of the present invention will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0017] Figure 1 This is a flowchart of an obstacle avoidance path generation method for an intelligent connected low-speed teaching vehicle according to the present invention. Figure 2 This is a flowchart of the obstacle avoidance algorithm logic for the obstacle avoidance path generation method of an intelligent connected low-speed teaching vehicle according to the present invention. Figure 3 This is a flowchart of the process for sensing static obstacles around a vehicle in this invention; Figure 4 This is a schematic diagram of the decision conditions upon which the generation of obstacle avoidance paths depends in this invention; Figure 5 This is a schematic diagram of the geometric principle of deflection angle generation in this invention. Detailed Implementation

[0018] The invention will now be described in more detail with reference to the accompanying drawings. While some embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the invention. It should be understood that the drawings and embodiments of the invention are for illustrative purposes only and are not intended to limit the scope of protection of the invention.

[0019] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0020] It should be noted that the concepts of "first" and "second" mentioned in this invention are only used to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0021] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0022] The names of messages or information exchanged between the various devices of this invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0023] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0024] like Figure 1 The diagram shows a flowchart of an obstacle avoidance path generation method for an intelligent connected low-speed teaching vehicle according to the present invention, which specifically includes the following steps: Step 101: Obtain the vehicle's location information, predetermined driving path information, and preset obstacle avoidance decision parameters, wherein the preset obstacle avoidance decision parameters include obstacle avoidance distance, safety factor, obstacle avoidance path threshold, detour distance threshold, and filtering factor. In some embodiments, the application scenario of the obstacle avoidance path generation method for an intelligent connected low-speed teaching vehicle according to the present invention is a teaching application scenario where the intelligent connected low-speed teaching vehicle generates an obstacle avoidance path when encountering static obstacles during straight-line autonomous driving tracking. The intelligent connected low-speed teaching vehicle refers to a low-speed autonomous driving vehicle used for autonomous driving teaching experiments in secondary vocational schools, higher vocational schools, or applied undergraduate programs. This vehicle has the following onboard systems: a combined navigation and positioning system (for obtaining vehicle position), a lidar perception system (for obtaining surrounding environmental information), an onboard decision control system (for executing the obstacle avoidance path generation algorithm), and a drive-by-wire execution system (including a drive-by-wire drive system, a brake-by-wire system, and a steering-by-wire system). In the present invention, during straight-line autonomous driving tracking, the vehicle automatically drives according to a predetermined driving path in a pre-generated tracking map, with a distance of 0.5 meters between each tracking point. In this application scenario, when a static obstacle appears in front of the vehicle, the vehicle needs to generate an obstacle avoidance path based on the perception results. The executing entity of the present invention is the onboard decision control system. The onboard decision control system refers to the computing unit installed on the intelligent connected low-speed teaching vehicle, used to execute the autonomous driving decision algorithm. The vehicle-mounted decision control system can acquire data through communication methods such as serial communication, CAN bus communication, or Ethernet communication. A tracking map is a two-dimensional coordinate map used for autonomous driving tracking control; this map records the coordinates of each tracking point along a predetermined driving path in planar coordinate form.

[0025] In some embodiments, such as Figure 2 As shown, a flowchart illustrating the obstacle avoidance algorithm logic of an obstacle avoidance path generation method for an intelligent connected low-speed teaching vehicle according to the present invention is presented, wherein... Figure 2The demonstration shows that: First, the vehicle uses a perception system such as LiDAR to detect static obstacles around it and acquire information about the surrounding environment. Then, the vehicle determines whether the static obstacle is in its path. This determination step has two possible outcomes: if the static obstacle is not in the path ("No"), the vehicle maintains normal driving and does not trigger obstacle avoidance; if the static obstacle is in the path ("Yes"), an obstacle avoidance event is triggered, and the vehicle enters the obstacle avoidance path generation phase. In the obstacle avoidance path generation phase, the vehicle regenerates a detour path based on the obstacle's size (including length and width) and its distance from the vehicle (distance between the vehicle and the obstacle). The vehicle then controls itself according to the generated detour path, driving along it to avoid the obstacle. During the detour, the vehicle continuously checks whether it has successfully avoided the obstacle: if it has not successfully avoided the obstacle ("No"), it continues driving along the current obstacle avoidance path; if it has successfully avoided the obstacle ("Yes"), it enters the regression path generation phase. During the path regression generation phase, the vehicle generates a path to return to the predetermined driving route, that is, gradually returning to the original driving route from the end of the obstacle avoidance path. Finally, the vehicle is controlled according to the generated regression path, causing the vehicle to travel along the regression path and eventually return to the predetermined driving route, completing the entire static obstacle avoidance process. The onboard decision control system obtains the vehicle's positioning information through the integrated navigation and positioning system. The integrated navigation and positioning system refers to a system that simultaneously uses the following two types of positioning devices for fusion positioning: a GPS receiver and an inertial navigation unit. The integrated navigation and positioning system outputs the vehicle's current position through a data fusion algorithm. The onboard decision control system receives the data output by the integrated navigation and positioning system through serial communication or Ethernet communication. For example, the integrated navigation and positioning system sends positioning data, i.e., the vehicle's positioning information, to the onboard decision control system every 100 milliseconds. Vehicle positioning information refers to the coordinate information describing the vehicle's current position in the tracking map. Vehicle positioning information includes the vehicle's lateral coordinates and longitudinal coordinates. The predetermined driving path information refers to the target path that the vehicle needs to follow during the autonomous driving tracking process. The predetermined driving path information consists of multiple tracking points. A tracking point is a discrete coordinate point on the predetermined driving path. Each tracking point contains two parameters: its lateral coordinates and its longitudinal coordinates. In this invention, the distance between two adjacent tracking points is 0.5 meters. The predetermined driving path information is pre-stored in the path file of the onboard decision control system. The onboard decision control system obtains the predetermined driving path information by reading the path file. After reading the file, the onboard decision control system stores the path points as the predetermined driving path information. The preset obstacle avoidance decision parameters refer to the set of parameters used to control the shape of the obstacle avoidance path when generating it. These preset obstacle avoidance decision parameters are configured by instructors in the system before the vehicle's autonomous driving begins.The configuration can be done in the following ways: inputting the parameters in the vehicle decision control system configuration interface or writing them in the parameter configuration file. Here, obstacle avoidance distance refers to the distance threshold used to trigger the obstacle avoidance function. Safety factor is a parameter used to adjust the angle between the obstacle avoidance path and the predetermined driving path. The larger the safety factor, the larger the angle of the generated obstacle avoidance path. Obstacle avoidance path threshold is a parameter used to determine the end point of the obstacle avoidance path. The obstacle avoidance path threshold is based on the longitudinal coordinates of the obstacle's center point. Detour distance threshold is a parameter used to calculate the proportion of the straight-line distance the vehicle needs to continue traveling parallel to the predetermined driving path after bypassing the obstacle. Filtering coefficient is a parameter used to adjust the spacing between tracking points in the obstacle avoidance path. Through the above methods, the vehicle decision control system obtains vehicle positioning information, predetermined driving path information, and preset obstacle avoidance decision parameters, providing basic data for subsequent steps to detect the existence of static obstacles on the predetermined driving path and generate obstacle avoidance paths.

[0026] Step 102: Detect whether there are static obstacles on the predetermined driving path; when a static obstacle is detected on the predetermined driving path, obtain the position information and size information of the static obstacle, and convert the position information of the static obstacle to the vehicle coordinate system; In some embodiments, such as Figure 3The diagram illustrates a flowchart of the process for perceiving static obstacles around a vehicle in this invention. Specifically, firstly, the vehicle uses a sensing system such as LiDAR to perceive its surrounding environment and acquire environmental information, including point cloud data. Then, based on the perceived environmental information, the vehicle determines whether a static obstacle exists on its driving route. This determination step has two branches: when no static obstacle exists on the driving route (i.e., the "No" branch), the vehicle maintains normal driving and does not trigger obstacle avoidance; when a static obstacle exists on the driving route (i.e., the "Yes" branch), the vehicle transmits the obstacle information to the decision-making mechanism. The transmitted obstacle information includes the obstacle's location information (such as the coordinates of the obstacle's center point) and size information (such as the obstacle's length and width), providing input data for subsequent obstacle avoidance path generation. Through this process, the vehicle completes the perception and identification of static obstacles and transmits the relevant data to the decision-making mechanism, laying the foundation for subsequent obstacle avoidance path generation. The onboard decision control system detects the presence of static obstacles on the predetermined driving path by receiving environmental information output from the LiDAR sensing system. Static obstacles refer to objects located in front of a vehicle during autonomous driving tracking that remain in a fixed position for a certain period of time. In the teaching application scenario of this invention, static obstacles may include traffic cones, roadblocks, teaching test boxes, or parked vehicles. Static obstacles have the following geometric attributes: obstacle position, obstacle length, and obstacle width, which are obtained by the LiDAR perception system. The vehicle's LiDAR perception system scans the environment around the vehicle using LiDAR and generates environmental point cloud data. Environmental point cloud data refers to a set of spatial coordinate points obtained by LiDAR scanning, where each point contains the following data: the point's lateral coordinate, the point's longitudinal coordinate, and the point's height coordinate. The onboard decision control system receives the environmental point cloud data output by the LiDAR perception system via Ethernet communication and then performs clustering processing on the environmental point cloud data to identify obstacles in the environment. Clustering processing refers to grouping point cloud data with similar spatial distances into the same point set; for example, Euclidean distance clustering algorithms can be used for clustering processing. After clustering processing, each point cloud set represents an obstacle target. For each identified point cloud set, the vehicle decision control system generates an obstacle bounding box. An obstacle bounding box is the smallest rectangular region that can enclose the point cloud set. The obstacle bounding box includes the coordinates of the obstacle's center point, the obstacle's length, and the obstacle's width. The obstacle length refers to the size of the bounding box in the vehicle's driving direction, and the obstacle width refers to the size of the bounding box in the vehicle's lateral direction. Based on this, the vehicle decision control system compares the center point coordinates of each obstacle bounding box with the predetermined driving path information to determine whether the obstacle is located on the predetermined driving path. This determination can be implemented as follows: First, a path detection region is established on the predetermined driving path.The path detection area refers to a rectangular detection area established around a predetermined driving path. The width of this area can be set to the vehicle width. For example, if the vehicle width is 1.2 meters, the path detection area width can be set to 1.5 meters. Subsequently, the onboard decision control system determines whether the center point of an obstacle is located within the path detection area. When the center point of the obstacle is within the path detection area, it is determined that a static obstacle exists on the predetermined driving path. When the center point of the obstacle is not located within the path detection area, it is determined that no static obstacle exists on the predetermined driving path. When a static obstacle is detected on the predetermined driving path, the onboard decision control system acquires the position information of the static obstacle. The position information of a static obstacle refers to the coordinate information describing the position of the static obstacle in the environmental coordinate system. The position information of a static obstacle includes the lateral coordinates and longitudinal coordinates of the obstacle's center point. The size information of a static obstacle refers to the data describing the geometric dimensions of the static obstacle. The size information of a static obstacle includes the obstacle's length and width. The obstacle position information output by the LiDAR perception system is the coordinates in the environmental coordinate system. To perform subsequent obstacle avoidance path calculations, the positional information of static obstacles needs to be converted into coordinates in the vehicle coordinate system. The environmental coordinate system is a two-dimensional plane coordinate system established with reference to the tracking map. In this system, the x-axis represents the lateral direction, and the y-axis represents the vehicle's driving direction. The vehicle coordinate system is a coordinate system established with the vehicle's current position as the origin. The vehicle coordinate system has the following definitions: Origin: current vehicle position; y-axis direction: current vehicle driving direction; x-axis direction: lateral direction of the vehicle. Based on this, the onboard decision control system transforms the obstacle coordinates from the environmental coordinate system to the vehicle coordinate system according to the vehicle's positioning information. Specifically, first, the obstacle coordinates in the environmental coordinate system and the vehicle positioning information are obtained. The obstacle coordinates in the environmental coordinate system include the lateral and longitudinal coordinates of the obstacle, and the vehicle positioning information includes the lateral and longitudinal coordinates of the vehicle in the environmental coordinate system. Then, using the vehicle's position in the environmental coordinate system as a reference point, the lateral coordinates of the obstacle in the environmental coordinate system are subtracted from the vehicle's lateral coordinates to obtain the obstacle's lateral coordinates in the vehicle coordinate system. Simultaneously, the longitudinal coordinates of the obstacle in the environmental coordinate system are subtracted from the longitudinal coordinates of the vehicle to obtain the longitudinal coordinates of the obstacle in the vehicle coordinate system. Through this coordinate transformation, the position data of the static obstacle in the vehicle coordinate system can be obtained. This position data can be used in subsequent steps to calculate the distance between the vehicle and the obstacle, determine the obstacle avoidance path angle, and generate obstacle avoidance path tracking points.

[0027] Step 103: Determine the distance information between the vehicle and the static obstacle based on the position information of the static obstacle in the vehicle coordinate system; In some embodiments, the onboard decision control system has acquired the position data of a static obstacle in the vehicle coordinate system. This position data includes the obstacle's lateral coordinates and longitudinal coordinates. The lateral coordinates refer to the numerical value of the static obstacle's position along the vehicle's lateral direction in the vehicle coordinate system. The longitudinal coordinates refer to the numerical value of the static obstacle's position along the vehicle's travel direction in the vehicle coordinate system. The origin of the vehicle coordinate system is the vehicle's current position; therefore, the vehicle's position coordinates in the vehicle coordinate system are (0,0). The distance information between the vehicle and the static obstacle refers to the spatial distance between the vehicle's current position and the static obstacle. In this embodiment, this distance is the Euclidean distance in the vehicle coordinate system. Euclidean distance refers to the straight-line distance between two coordinate points in a two-dimensional plane. Specifically, the distance between the vehicle and the static obstacle is the square root of the sum of the squares of the lateral and longitudinal coordinates of the static obstacle. For example, when the lateral coordinate of a static obstacle in the vehicle coordinate system is 1.2 meters and the longitudinal coordinate is 4.5 meters, the squares of the lateral and longitudinal coordinates are first calculated separately, yielding a lateral distance squared of 1.44 and a longitudinal distance squared of 20.25. These are then added together to obtain 21.69, and the square root of this value is taken to obtain a distance of approximately 4.66 meters between the vehicle and the static obstacle. After completing the above calculations, the onboard decision control system generates distance information between the vehicle and the static obstacle. This distance information can include the spatial distance between the vehicle and the static obstacle, the lateral distance of the obstacle relative to the vehicle, and the longitudinal distance of the obstacle relative to the vehicle. This distance information can be used as input parameters for subsequent steps to determine whether the vehicle needs to perform obstacle avoidance maneuvers, and to calculate the obstacle avoidance path angle and generate the obstacle avoidance path.

[0028] Step 104: Determine the obstacle avoidance starting position based on the obstacle avoidance distance and distance information; Among these steps, determining the obstacle avoidance starting position based on the obstacle avoidance distance and distance information includes: When the distance information is less than the obstacle avoidance distance, the obstacle avoidance condition is triggered, and the current vehicle position is used as the starting tracking point of the obstacle avoidance path.

[0029] In some embodiments, the vehicle-mounted decision control system first reads the distance information and compares it with the obstacle avoidance distance. When the distance information is less than the obstacle avoidance distance, the obstacle avoidance condition is deemed met. At this time, the vehicle-mounted decision control system internally sets an obstacle avoidance trigger flag, instructing the obstacle avoidance path generation module to begin generating an obstacle avoidance path. If the distance information is greater than or equal to the obstacle avoidance distance, the obstacle avoidance condition is not met, and the vehicle-mounted decision control system does not execute the obstacle avoidance action. When the obstacle avoidance condition is triggered, the vehicle-mounted decision control system uses the current vehicle position as the starting tracking point of the obstacle avoidance path. The current vehicle position is defined as the origin in the vehicle coordinate system, with a horizontal coordinate of 0 and a vertical coordinate of 0. The vehicle-mounted decision control system creates the first tracking point object and sets its coordinates to the current vehicle position. The path generation for subsequent obstacle avoidance segments will start from this initial tracking point and gradually generate various tracking points on the obstacle avoidance path based on the longitudinal step distance and deflection angle. For example, when the obstacle's horizontal coordinate in the vehicle coordinate system is 1.0 meter and its vertical coordinate is 2.5 meters, and the obstacle avoidance distance is 3 meters, the actual distance between the vehicle and the obstacle is approximately 2.69 meters. If this distance is less than the obstacle avoidance distance, the obstacle avoidance condition is met. The vehicle decision control system uses the current vehicle position (0,0) as the starting tracking point for the obstacle avoidance path and then generates the obstacle avoidance path segment. Through the above operations, the vehicle decision control system determines the starting position for obstacle avoidance path generation, providing basic data and starting coordinates for subsequent calculations of obstacle avoidance path deflection angle, longitudinal step, and lateral coordinate changes.

[0030] Step 105: Calculate the deflection angle of the obstacle avoidance path relative to the predetermined driving path based on the size information, distance information, and safety factor of the static obstacle. In some embodiments, such as Figure 4The diagram illustrates the decision conditions upon which obstacle avoidance paths are generated in this invention, specifically showing the input data and decision parameters required for path generation. The diagram shows that the decision conditions for generating obstacle avoidance paths consist of two main parts: coordinate system transformation and obstacle avoidance decision conditions. In the coordinate system transformation part, the vehicle needs to transform the perceived obstacle information into the vehicle coordinate system. This part includes the following two key input data: Distance between the obstacle and the vehicle: This refers to the positional distance of the static obstacle relative to the vehicle, including the lateral and longitudinal distances of the obstacle in the vehicle coordinate system. This is used to determine the spatial relationship between the vehicle and the obstacle, providing a basis for calculating the obstacle avoidance path deflection angle and determining the obstacle avoidance trigger timing. Obstacle volume: This refers to the geometric dimensions of the static obstacle, including the obstacle length and width. This is used to determine the lateral offset and longitudinal detour distance required for obstacle avoidance path generation. The result of the above coordinate system transformation is to convert the obstacle's positional information in the environmental coordinate system into coordinates in the vehicle coordinate system, enabling subsequent obstacle avoidance path calculations to be performed in a coordinate system with the vehicle as the origin. In the obstacle avoidance decision-making section, the vehicle needs to generate an obstacle avoidance path based on the following five preset obstacle avoidance decision parameters: safety factor, detour distance threshold, filtering factor, obstacle avoidance distance, and obstacle avoidance path threshold. By combining the obstacle distance and volume information obtained through the coordinate system transformation with the five obstacle avoidance decision parameters, the vehicle can dynamically generate obstacle avoidance paths adapted to different obstacle conditions and teaching needs. The deflection angle refers to the angle between the obstacle avoidance path and the predetermined driving path, used to control the magnitude of the obstacle avoidance path's offset in the vehicle's lateral direction. First, the onboard decision control system acquires the size information of the static obstacle, including the obstacle's length and width. Then, the onboard decision control system acquires distance information, i.e., the distance from the vehicle's current position to the obstacle's key points. In this embodiment of the invention, the longitudinal and lateral distances from the obstacle's left front or lower left corner point to the vehicle's current position can be used to calculate the straight-line distance between the vehicle and the obstacle. The straight-line distance is calculated using a two-dimensional Euclidean distance formula, i.e., the square root of the square of the lateral coordinate plus the square of the longitudinal coordinate, yielding the reference distance between the vehicle and the obstacle. Subsequently, the onboard decision control system calculates the deflection angle of the obstacle avoidance path based on the safety factor and the obstacle width. The safety factor is a parameter configured by the instructor before the vehicle begins autonomous driving. It is used to adjust the angle between the obstacle avoidance path and the predetermined driving path. In this implementation, the vehicle's current position is used as the starting point for generating the obstacle avoidance path. Adjusting the safety factor changes the angle between the obstacle avoidance path and the predetermined driving path. A larger safety factor results in a larger angle between the obstacle avoidance path and the predetermined driving path, thus maintaining a greater lateral safety distance when the vehicle avoids obstacles. A smaller safety factor results in a smaller angle, making the obstacle avoidance path closer to the predetermined driving path. In the specific calculation process, the onboard decision control system first transforms the coordinates of the obstacle to the vehicle coordinate system based on the vehicle's current position information.Subsequently, a basic angle is calculated based on the obstacle's width and the distance from the vehicle's current position to the key point at the left front or lower left corner of the obstacle. This basic angle can be expressed as the geometric angle formed by the line connecting the vehicle's current position to the obstacle's center point and the line connecting the vehicle's current position to the key point at the left front of the obstacle. After obtaining the basic angle, it is multiplied by a safety factor to obtain the final deflection angle used to generate the obstacle avoidance path. In specific implementation, the formula for calculating the deflection angle is:

[0031] in, This represents the arcsine function, with the output angle unit being degrees or radians. The units for obstacle width and distance remain consistent, for example, meters. After calculation, the vehicle-mounted decision control system obtains the deflection angle value between the obstacle avoidance path and the predetermined driving path. For example... Figure 5 The diagram shows the geometric principle of deflection angle generation in this invention. It assumes that the angle between the line segment representing the distance from the current vehicle to the center point of the obstacle and the line segment representing the distance from the current vehicle to the left front of the obstacle is 10°. 。 If the safety factor is 1.0, the included angle of the obstacle avoidance path is 10°. 。 (i.e., 10 × 1.0). If the safety factor is 1.5, the included angle of the obstacle avoidance path is 15°. 。 (10×1.5).

[0032] Step 106: Generate an obstacle avoidance path around static obstacles based on the obstacle avoidance starting position, deflection angle, obstacle avoidance path threshold, detour distance threshold, and filtering coefficient. Generating obstacle avoidance paths around static obstacles includes: The end point of the obstacle avoidance path is determined based on the obstacle avoidance path threshold.

[0033] In some embodiments, during the generation of an obstacle avoidance path around a static obstacle, the end tracking point of the obstacle avoidance path must first be determined based on an obstacle avoidance path threshold. The end tracking point refers to the coordinates of the position where the obstacle avoidance segment terminates; that is, after the vehicle travels along the deflection direction to this position, the obstacle avoidance segment ends, and the vehicle enters the path planning stage for subsequent detours or returning to the predetermined driving path. In this embodiment, the obstacle avoidance path threshold is used to control the length of the obstacle avoidance path in the vehicle's driving direction. To determine the location of the end tracking point, the onboard decision control system first obtains the longitudinal coordinates of the center point of the static obstacle in the vehicle coordinate system. The center point of the static obstacle is the center point of the bounding box obtained by clustering the obstacle point cloud through the perception system; its longitudinal coordinates represent the distance of the obstacle in the vehicle's driving direction. Subsequently, the longitudinal coordinates of the end tracking point are calculated based on the obstacle avoidance path threshold. The obstacle avoidance distance refers to the distance from the vehicle's current position to the center point of the obstacle in the vehicle's driving direction; this distance can be directly determined by the vehicle coordinates and the longitudinal coordinates of the obstacle's center point. In one implementation, the longitudinal coordinate of the end tracking point can be determined as follows: Longitudinal coordinate of end tracking point = Longitudinal coordinate of obstacle center point × Obstacle avoidance path threshold. Through this calculation, obstacle avoidance segments of different lengths can be generated based on different obstacle avoidance path thresholds. When the obstacle avoidance path threshold is 1.0, the longitudinal coordinate of the end tracking point is equal to the longitudinal coordinate of the obstacle center point. In this case, the obstacle avoidance segment ends when the vehicle travels along the obstacle avoidance path to the same longitudinal position as the obstacle center point. In this situation, the longitudinal endpoint of the vehicle's obstacle avoidance path is at the same longitudinal position as the obstacle center point. When the obstacle avoidance path threshold is less than 1.0, the absolute value of the longitudinal coordinate of the end tracking point is less than the absolute value of the longitudinal coordinate of the obstacle center point. This means that the obstacle avoidance segment ends prematurely before reaching the obstacle center point, at which point the vehicle will begin to detour or return to the predetermined path at a position closer to itself, thus generating a shorter obstacle avoidance path. When the obstacle avoidance path threshold is greater than 1.0, the absolute value of the longitudinal coordinate of the end tracking point is greater than the absolute value of the longitudinal coordinate of the obstacle center point. In this scenario, the vehicle continues to travel along the deflection direction beyond the obstacle's center point during obstacle avoidance, increasing the length of the obstacle avoidance segment and thus expanding the safety margin when navigating around the obstacle. Through this method, the onboard decision control system can flexibly adjust the obstacle avoidance path length based on the obstacle avoidance path threshold and determine the end tracking point of the obstacle avoidance path, providing termination coordinates for the subsequent generation of obstacle avoidance and detour segments.

[0034] The dimensional information of static obstacles includes the obstacle's length and width, as well as... Generating obstacle avoidance paths around static obstacles also includes: Calculate the straight-line detour distance based on the obstacle length and the detour distance threshold; Starting from the end of the tracking point, generate a detour segment with a length equal to the straight-line distance of the detour and parallel to the predetermined driving path.

[0035] In some embodiments, the process of generating an obstacle avoidance path around a static obstacle further includes calculating the straight-line detour distance based on the length of the static obstacle and a detour distance threshold. The detour distance threshold is a pre-set parameter used to determine the straight-line distance the vehicle must travel along the predetermined driving path after bypassing the obstacle. Its function is to ensure that the vehicle gradually returns to the predetermined driving path after completely passing the obstacle. In a specific implementation, the onboard decision control system first obtains the length value of the static obstacle and then reads the preset detour distance threshold. The detour straight-line distance is calculated as follows: Detour straight-line distance = Obstacle length × Detour distance threshold. Through the above calculation, the distance the vehicle needs to continue parallel driving after bypassing the obstacle can be obtained. When the detour distance threshold is large, the calculated detour straight-line distance is long, and the vehicle needs to continue traveling a longer distance outside the obstacle avoidance path to ensure that the vehicle completely bypasses the obstacle and maintains a large safety margin; when the detour distance threshold is small, the detour straight-line distance is short, and the vehicle begins to return to the predetermined driving path more quickly after bypassing the obstacle. For example, in one specific embodiment, when the obstacle length is 2 meters and the detour distance threshold is 1.5, the detour straight-line distance is: detour straight-line distance = 2 × 1.5 = 3 meters. After bypassing the obstacle, the vehicle needs to continue traveling 3 meters in a parallel direction. After obtaining the detour straight-line distance, the vehicle decision control system generates a detour segment starting from the end tracking point. The end tracking point is the coordinate of the endpoint of the obstacle avoidance segment determined in the aforementioned steps, and its coordinates include longitudinal and lateral coordinates. The detour segment refers to a straight path that the vehicle maintains parallel to the predetermined driving path after bypassing the obstacle. Specifically, the detour segment is generated by using the end tracking point as the starting point of the detour segment, keeping the lateral coordinate unchanged in the vehicle coordinate system, thus maintaining a constant offset distance from the predetermined driving path in the lateral direction. Simultaneously, the longitudinal coordinate is gradually increased along the predetermined driving path until the cumulative increase in the longitudinal coordinate reaches the calculated detour straight-line distance. This method generates a detour segment with a length equal to the detour straight-line distance and parallel to the predetermined driving path. In terms of path representation, this detour segment also consists of multiple consecutive tracking points. The longitudinal interval between each tracking point is determined by the longitudinal step distance set by the system, thus forming a discrete straight-line path. During autonomous driving control, the vehicle tracks the trajectory according to this sequence of tracking points, thereby achieving the process of continuing parallel driving after bypassing static obstacles. In this way, the onboard decision control system can dynamically calculate the detour straight-line distance based on the length of the obstacle and the detour distance threshold, and generate a detour segment parallel to the predetermined driving path at the end position of the obstacle avoidance path, thereby ensuring that the vehicle has sufficient driving distance to complete a safe detour after passing the obstacle.

[0036] Generating obstacle avoidance paths around static obstacles also includes: The longitudinal step distance between tracking points is determined based on the filtering coefficient and the preset default distance.

[0037] In some embodiments, generating an obstacle avoidance path around static obstacles further includes determining the longitudinal step distance between tracking points based on a filtering coefficient and a preset default distance. The longitudinal step distance refers to the distance between two adjacent tracking points in the vehicle's driving direction during obstacle avoidance path generation; this distance controls the generation density of tracking points in the obstacle avoidance path. In this embodiment, the obstacle avoidance path consists of multiple discrete tracking points, each containing lateral and longitudinal coordinates. During autonomous driving, the vehicle sequentially tracks these tracking points; therefore, the distance between tracking points directly affects the smoothness of the path and the number of path points. Before determining the longitudinal step distance, the onboard decision control system first reads the preset default distance in the system. The default distance is a pre-set base step distance used as a reference value for the interval between tracking points. In this embodiment, the default distance between tracking points is 0.5 meters, meaning that without additional adjustment, the longitudinal interval between two adjacent tracking points is 0.5 meters. Subsequently, the onboard decision control system obtains the filtering coefficient parameters. The filtering coefficient is a parameter used to adjust the number of tracking points in the obstacle avoidance path. Its function is to change the spacing between tracking points, thereby controlling the density of tracking points in the path. A larger filtering coefficient results in a larger distance between tracking points, thus generating fewer tracking points; a smaller filtering coefficient results in a smaller distance between tracking points, generating a relatively larger number of tracking points. In specific implementations, the onboard decision control system calculates the longitudinal step distance between tracking points based on the filtering coefficient and the default distance. The calculation method is as follows: Longitudinal step distance = Filtering coefficient × Default distance. This calculation yields the longitudinal interval distance between adjacent tracking points during obstacle avoidance path generation. For example, in one specific embodiment, when the filtering coefficient is set to 1, the distance between tracking points after obstacle avoidance path generation remains at the default distance, i.e., 0.5 meters. In this case, a new tracking point is generated every 0.5 meters, resulting in a relatively large number of tracking points in the generated obstacle avoidance path and a more detailed path description. In another embodiment, when the filtering coefficient is set to 2, the vehicle decision control system multiplies the default distance of 0.5 meters by the filtering coefficient 2 to obtain a longitudinal step distance of 1.0 meter, i.e., longitudinal step distance = 2 × 0.5 = 1.0 meter. At this point, after generating the obstacle avoidance path, the longitudinal interval between adjacent tracking points becomes 1.0 meter. Because the distance between tracking points increases, the number of tracking points generated within the same path length decreases, thereby reducing the number of path points. After completing the above calculation, the vehicle decision control system uses the obtained longitudinal step distance as the stepping parameter in the path generation process. When subsequently generating obstacle avoidance and detour sections, this longitudinal step distance is used as the interval for longitudinal coordinate progression, thereby sequentially generating multiple tracking points and forming a complete obstacle avoidance path. In this way, the number and distribution density of tracking points in the obstacle avoidance path can be flexibly controlled according to different filtering coefficients.

[0038] Generating obstacle avoidance paths around static obstacles also includes: Between the starting position and the ending tracking point of obstacle avoidance, the longitudinal coordinate is maintained by a longitudinal step distance, and the change in the lateral coordinate is determined based on the longitudinal step distance and the deflection angle to generate the obstacle avoidance segment.

[0039] The change in the lateral coordinate is determined based on the longitudinal step distance and the deflection angle, including: The lateral coordinate increment is determined based on the sine value of the deflection angle and the longitudinal step distance; Based on the vehicle's obstacle avoidance direction, the current lateral coordinate is offset using the lateral coordinate increment.

[0040] In some embodiments, generating an obstacle avoidance path around static obstacles further includes generating an obstacle avoidance segment between the starting position and the ending tracking point of the obstacle avoidance. In this embodiment, the default distance between two adjacent tracking points in the predetermined driving path is 0.5 meters. Therefore, when generating the obstacle avoidance segment, the longitudinal coordinate still progresses stepwise at 0.5-meter intervals to maintain consistency between the obstacle avoidance path and the predetermined path in terms of tracking point spacing. To ensure that the actual spatial distance between each tracking point in the generated obstacle avoidance path remains approximately 0.5 meters, while the longitudinal coordinate progresses at 0.5-meter intervals, it is also necessary to calculate the corresponding lateral coordinate change based on the deflection angle of the obstacle avoidance path. Specifically, the onboard decision control system calculates the increment of the lateral coordinate based on the sine value of the deflection angle and the longitudinal step distance. The calculation method is: lateral coordinate increment = sin(deflection angle) × longitudinal step distance. When the longitudinal step distance is 0.5 meters, the change in the lateral coordinate each time is sin(deflection angle) × 0.5. When generating a new tracking point, the longitudinal coordinate increases by 0.5 meters from the previous tracking point, while the lateral coordinate is superimposed or subtracted based on the obstacle avoidance direction. When the vehicle chooses to circumvent an obstacle to the left, the lateral coordinate increment is added to the current lateral coordinate; when the vehicle chooses to circumvent an obstacle to the right, the lateral coordinate increment is subtracted from the current lateral coordinate. By continuously executing the above coordinate calculation process, multiple tracking points can be gradually generated, thus forming an obstacle avoidance segment path that forms a certain angle with the predetermined driving path. An obstacle avoidance segment refers to a path formed by the vehicle starting from the obstacle avoidance starting position, gradually shifting to one side according to a pre-calculated deflection angle, and finally reaching the ending tracking point. An obstacle avoidance segment consists of multiple consecutive tracking points, each containing both longitudinal and lateral coordinates. During autonomous driving, the vehicle tracks the trajectory sequentially according to these tracking points to complete the driving process of bypassing static obstacles. When generating an obstacle avoidance segment, the onboard decision control system first uses the obstacle avoidance starting position as the first tracking point. The obstacle avoidance starting position is the point where the vehicle begins to deviate from the predetermined driving path. The coordinates of this position include initial longitudinal and initial lateral coordinates. Subsequently, between the obstacle avoidance starting position and the end tracking point, the longitudinal coordinates are progressively increased according to the previously determined longitudinal step distance to generate new tracking points. Specifically, based on the longitudinal coordinates of the current tracking point, the longitudinal step distance is used as the increment, making the longitudinal coordinates of the next tracking point equal to the current longitudinal coordinate plus the longitudinal step distance. By repeating this operation, multiple progressively increasing longitudinal coordinates can be obtained until the longitudinal coordinates reach or approach the longitudinal coordinates of the end tracking point. After determining the longitudinal coordinates of each new tracking point, the change in lateral coordinates also needs to be determined based on the longitudinal step distance and deflection angle, thus causing the path to gradually shift towards the obstacle avoidance direction. In one specific implementation, the process of determining the change in lateral coordinates based on the longitudinal step distance and deflection angle includes the following steps: First, the lateral coordinate increment is calculated based on the sine value of the deflection angle and the longitudinal step distance.The lateral coordinate increment represents the offset between two adjacent tracking points in the lateral direction of the vehicle. It is calculated as follows: Lateral coordinate increment = Longitudinal step distance × sin(deflection angle), where sin represents a sine function, the deflection angle is the angle between the obstacle avoidance path and the predetermined driving path, and the longitudinal step distance is the longitudinal interval between adjacent tracking points. This calculation yields the lateral offset corresponding to the current longitudinal step distance. Subsequently, the current lateral coordinate is offset according to the vehicle's obstacle avoidance direction. The obstacle avoidance direction refers to the detour direction chosen by the vehicle when bypassing an obstacle, typically including either left or right detours. When the vehicle chooses to detour left, the lateral coordinate of the next tracking point equals the current lateral coordinate plus the calculated lateral coordinate increment; when the vehicle chooses to detour right, the lateral coordinate of the next tracking point equals the current lateral coordinate minus the calculated lateral coordinate increment. This calculation yields the lateral coordinate of the next tracking point. Then, this lateral coordinate is combined with the corresponding longitudinal coordinate to generate the new tracking point coordinates. The vehicle-mounted decision control system performs calculations cyclically as described above: each time, the longitudinal coordinate is increased by a longitudinal step distance, while the corresponding lateral coordinate increment is calculated based on the deflection angle, and the lateral coordinate is offset according to the obstacle avoidance direction, thus gradually generating multiple tracking points. As tracking points are continuously generated, the path gradually tilts towards the obstacle avoidance direction and eventually extends to the end tracking point. In this way, a continuous obstacle avoidance path segment can be generated between the starting position and the end tracking point. This obstacle avoidance segment consists of multiple tracking points that progress at fixed longitudinal step distances and gradually shift laterally, allowing the vehicle to smoothly bypass static obstacles along this path.

[0041] Step 107: After the vehicle bypasses the static obstacle, a regression path is generated to return to the predetermined driving path.

[0042] The process of generating a regression path that returns to the predetermined driving path includes: Using the end point of the obstacle avoidance path as the starting point, the longitudinal coordinate is maintained at a preset step distance, and the regression change of the lateral coordinate is determined according to the preset step distance and deflection angle to generate a regression segment. The lateral coordinate change direction of the regression segment is opposite to that of the obstacle avoidance segment.

[0043] In some embodiments, after the vehicle completes its detour around a static obstacle, a return path needs to be generated to return to the predetermined driving path. The return path refers to a segment of the path from the end of the obstacle avoidance path back to the predetermined driving path after the vehicle has bypassed the obstacle, ultimately returning to the predetermined driving path. The return path consists of multiple consecutive tracking points, each including lateral and longitudinal coordinates. During autonomous driving, the vehicle sequentially tracks these tracking points to gradually return to the predetermined driving path. When generating the return path, the onboard decision control system first uses the end tracking point of the obstacle avoidance path as the starting point of the return path. The end tracking point of the obstacle avoidance path is the coordinate of the path endpoint determined after generating the obstacle avoidance path and the detour segment in the aforementioned steps; this coordinate includes longitudinal and lateral coordinates. Subsequently, the return segment is generated starting from this end tracking point. During the generation of the return segment, the onboard decision control system maintains the longitudinal coordinates progressively according to a preset step distance. The preset step distance refers to the distance between two adjacent tracking points in the vehicle's driving direction; in one embodiment, this step distance is 0.5 meters. The specific implementation is as follows: Each time a new tracking point is generated, the longitudinal coordinate of the current tracking point is increased or decreased by 0.5 meters to form a new longitudinal coordinate. To maintain consistent distances between tracking points in the regression path, while the longitudinal coordinate increases by 0.5 meters, the regression change in the lateral coordinate also needs to be calculated based on the preset step distance and deflection angle. The deflection angle is the angle between the obstacle avoidance path calculated in the previous steps and the predetermined driving path. In one specific implementation, the regression change in the lateral coordinate is determined as follows: First, the sine value of the deflection angle is calculated, and then this sine value is multiplied by the preset step distance to obtain the lateral coordinate change. The calculation formula is as follows: Lateral coordinate change = sin(deflection angle) × preset step distance. When the preset step distance is 0.5 meters, the lateral coordinate change is equal to the sine value of the deflection angle multiplied by 0.5. Subsequently, the lateral coordinate is offset according to the vehicle's regression direction. The vehicle's regression direction refers to the direction of movement of the vehicle gradually returning from the obstacle avoidance path to the predetermined driving path. The direction of lateral coordinate change in the regression path must be opposite to the direction of lateral coordinate change in the obstacle avoidance segment. For example, when the vehicle avoids an obstacle by moving to the left, the lateral coordinate in the obstacle avoidance segment gradually increases to the left; when generating the regression path, it needs to return to the predetermined driving path to the right. Therefore, when generating a new tracking point, the lateral coordinate change is subtracted from the current lateral coordinate. Conversely, when the vehicle avoids an obstacle by moving to the right, the lateral coordinate in the obstacle avoidance segment gradually decreases to the right; when generating the regression path, it needs to return to the predetermined driving path to the left. Therefore, when generating a new tracking point, the lateral coordinate change is added to the current lateral coordinate.In the above manner, when generating each new tracking point, the onboard decision control system performs the following calculation process: First, it increases the longitudinal coordinate of the current tracking point by a preset step distance to obtain a new longitudinal coordinate; then, it calculates the change in lateral coordinate based on the deflection angle; finally, it adds this change in lateral coordinate to the current lateral coordinate or subtracts it from the current lateral coordinate based on the regression direction, thus obtaining a new lateral coordinate. Combining the calculated lateral and longitudinal coordinates generates a new tracking point. By continuously repeating the above calculation process, the vehicle's path gradually shifts towards the predetermined driving path until the lateral coordinate gradually approaches the lateral coordinate of the predetermined driving path, thereby completing the process of generating the regression path.

[0044] In these embodiments, addressing the need for path planning and obstacle avoidance training in autonomous driving teaching scenarios using intelligent connected low-speed training vehicles, when static obstacles appear on the vehicle's predetermined driving path, obstacle avoidance-related decision parameters are set. This allows the vehicle to dynamically generate an obstacle avoidance path based on the obstacle's size and the distance between the vehicle and the obstacle, enabling the vehicle to complete the autonomous driving process of bypassing the obstacle and ultimately returning to the predetermined driving path. Furthermore, the generation of the obstacle avoidance path does not rely on SLAM maps or high-precision maps, nor on complex path planning algorithms. Instead, by setting decision parameters such as safety factor, detour distance threshold, filtering factor, obstacle avoidance distance, and obstacle avoidance path threshold, the obstacle avoidance path is dynamically generated based on the positional relationship between the vehicle and the static obstacle. This allows the vehicle to achieve obstacle avoidance functionality for static obstacles under relatively simple system conditions. Meanwhile, by adjusting the above decision parameters, various obstacle avoidance paths can be generated, enabling instructors to design different autonomous driving test cases based on vehicle speed, obstacle size, and testing requirements. By comparing and analyzing the actual obstacle avoidance results of the vehicle under different parameter settings, the perception information processing and decision-making process of the autonomous driving system during obstacle avoidance can be more intuitively demonstrated, which is conducive to the teaching and experimentation of autonomous driving-related technologies.

[0045] The above description is merely a selection of preferred embodiments of the present invention and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention is not limited to specific combinations of the above-described technical features, but also includes other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in this invention.

Claims

1. A method for generating obstacle avoidance paths for an intelligent connected low-speed teaching vehicle, characterized in that, include: The vehicle's location information, predetermined driving path information, and preset obstacle avoidance decision parameters are obtained. The preset obstacle avoidance decision parameters include obstacle avoidance distance, safety factor, obstacle avoidance path threshold, detour distance threshold, and filtering factor. Detect whether there are static obstacles on the predetermined driving path; when a static obstacle is detected on the predetermined driving path, obtain the position information and size information of the static obstacle, and convert the position information of the static obstacle to the vehicle coordinate system; Based on the position information of the static obstacle in the vehicle coordinate system, determine the distance information between the vehicle and the static obstacle; The obstacle avoidance starting position is determined based on the obstacle avoidance distance and the distance information. Based on the size information of the static obstacle, the distance information, and the safety factor, calculate the deflection angle of the obstacle avoidance path relative to the predetermined driving path; Based on the obstacle avoidance starting position, the deflection angle, the obstacle avoidance path threshold, the detour distance threshold, and the filtering coefficient, an obstacle avoidance path is generated to bypass the static obstacle; After the vehicle bypasses the static obstacle, a return path is generated to return to the predetermined driving path.

2. The obstacle avoidance path generation method for an intelligent connected low-speed teaching vehicle according to claim 1, characterized in that, Determining the obstacle avoidance starting position based on the obstacle avoidance distance and the distance information includes: When the distance information is less than the obstacle avoidance distance, the obstacle avoidance condition is triggered, and the current vehicle position is used as the starting tracking point of the obstacle avoidance path.

3. The obstacle avoidance path generation method for an intelligent connected low-speed teaching vehicle according to claim 1, characterized in that, The generation of an obstacle avoidance path that bypasses the static obstacle includes: The end point of the obstacle avoidance path is determined based on the obstacle avoidance path threshold.

4. The obstacle avoidance path generation method for an intelligent connected low-speed teaching vehicle according to claim 3, characterized in that, The size information of the static obstacle includes the obstacle's length and width.

5. The obstacle avoidance path generation method for an intelligent connected low-speed teaching vehicle according to claim 4, characterized in that, The generation of an obstacle avoidance path that bypasses the static obstacle further includes: Calculate the straight-line detour distance based on the obstacle length and the detour distance threshold; Starting from the end tracking point, a detour segment with a length equal to the detour straight-line distance and parallel to the predetermined driving path is generated.

6. The obstacle avoidance path generation method for an intelligent connected low-speed teaching vehicle according to claim 5, characterized in that, The generation of an obstacle avoidance path that bypasses the static obstacle further includes: Based on the filtering coefficient and the preset default distance, the longitudinal step distance between tracking points is determined, wherein the preset default distance is the distance between adjacent tracking points in a predetermined driving path.

7. The obstacle avoidance path generation method for an intelligent connected low-speed teaching vehicle according to claim 6, characterized in that, The generation of an obstacle avoidance path that bypasses the static obstacle further includes: Between the obstacle avoidance starting position and the ending tracking point, the longitudinal coordinate is maintained to advance according to the longitudinal step distance, and the change in the lateral coordinate is determined according to the longitudinal step distance and the deflection angle to generate the obstacle avoidance segment.

8. The obstacle avoidance path generation method for an intelligent connected low-speed teaching vehicle according to claim 7, characterized in that, The step of determining the change in lateral coordinates based on the longitudinal step distance and the deflection angle includes: Based on the sine value of the deflection angle and the longitudinal step distance, the lateral coordinate increment is determined; Based on the vehicle's obstacle avoidance direction, the current lateral coordinate is offset using the lateral coordinate increment.

9. The obstacle avoidance path generation method for an intelligent connected low-speed teaching vehicle according to claim 8, characterized in that, The generation of the regression path to the predetermined driving path includes: Using the end tracking point of the obstacle avoidance path as the starting point, the longitudinal coordinate is maintained at a preset step distance, and the regression change of the lateral coordinate is determined according to the preset step distance and the deflection angle to generate a regression segment, wherein the lateral coordinate change direction of the regression segment is opposite to the lateral coordinate change direction of the obstacle avoidance segment.

Images