Intelligent automatic driving method and system for IGV based on multi-mode navigation

Abstract

The application provides an IGV intelligent automatic driving method and system based on multi-mode navigation, which comprises the following steps: collecting first perception data, second perception data and third perception data of an intelligent guided vehicle (IGV) based on a laser radar sensor, a visual sensor and an inertial measurement sensor respectively; performing validity detection on the first perception data, the second perception data and the third perception data to obtain validity detection results of the respective perception data; determining effective perception data combinations based on the validity detection results of the respective perception data as input information of a current navigation mode to generate a candidate driving path; optimizing the candidate driving path based on a local environment obstacle map constructed based on the first perception data to obtain an optimal driving path; and controlling the IGV to automatically drive based on the optimal driving path. The application realizes reliable automatic driving of the IGV in a complex environment.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to an IGV intelligent autonomous driving method and system based on multi-mode navigation. Background Technology

[0002] In the field of industrial automation, intelligent guided vehicles (IGVs) are key equipment for material handling and production process automation, and their automatic driving control technology directly affects production efficiency and operational safety.

[0003] Existing intelligent autonomous driving control methods for IGVs facing malfunctions typically rely on a single perception system for navigation and control, such as using only LiDAR or visual sensors to acquire environmental information. When this single perception system malfunctions (e.g., the sensor is affected by dust or light interference, leading to abnormal data, or a sudden hardware failure), the IGV will be unable to locate and plan its path properly due to the interruption of environmental perception, which can lead to problems such as stalling, deviation from the planned route, or even collisions. This seriously affects the continuity and safety of industrial production, making it impossible for IGVs to achieve autonomous driving in complex environments. Summary of the Invention

[0004] This invention provides an IGV intelligent autonomous driving method and system based on multi-mode navigation, which can realize reliable autonomous driving of intelligent guided vehicles in complex environments.

[0005] In a first aspect, the present invention provides an IGV intelligent autonomous driving method based on multi-mode navigation, comprising:

[0006] Based on lidar sensors, vision sensors, and inertial measurement sensors, the first perception data, second perception data, and third perception data of the intelligent guided vehicle are collected respectively.

[0007] The validity of the first sensing data, the second sensing data, and the third sensing data is checked to obtain the validity check results of each sensing data.

[0008] The valid perception data determined based on the validity detection results of each perception data is used as the input information for the current navigation mode to generate candidate driving paths;

[0009] The candidate driving path is optimized based on the local environmental obstacle map constructed from the first perception data to obtain the optimal driving path.

[0010] The intelligent guided vehicle drives automatically based on the optimal driving path.

[0011] In a second aspect, the present invention also provides an IGV intelligent autonomous driving system based on multi-mode navigation, applied to the IGV intelligent autonomous driving method based on multi-mode navigation as described in the first aspect; the IGV intelligent autonomous driving system based on multi-mode navigation includes:

[0012] The data acquisition module is used to collect the first perception data, second perception data and third perception data of the intelligent guided vehicle based on the lidar sensor, vision sensor and inertial measurement sensor respectively;

[0013] The data detection module is used to perform validity detection on the first sensing data, the second sensing data, and the third sensing data, and obtain the validity detection results of each sensing data.

[0014] The driving route generation module is used to combine the valid perception data determined by the validity detection results of each perception data into the input information of the current navigation mode to generate candidate driving routes;

[0015] The driving path optimization module is used to optimize the candidate driving path based on the local environmental obstacle map constructed from the first perception data to obtain the optimal driving path.

[0016] The driving control module is used to control the intelligent guided vehicle to drive automatically based on the optimal driving path.

[0017] Thirdly, the present invention also provides an electronic device, comprising: a memory for storing computer software programs; and a processor for reading and executing the computer software programs, thereby realizing the IGV intelligent autonomous driving method based on multi-mode navigation as described above.

[0018] Fourthly, the present invention also provides a non-transitory computer-readable storage medium storing a computer software program, which, when executed by a processor, implements the IGV intelligent autonomous driving method based on multi-mode navigation as described above.

[0019] Fifthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the IGV intelligent autonomous driving method based on multi-mode navigation as described above.

[0020] The IGV intelligent autonomous driving method based on multi-mode navigation provided in this invention performs validity detection on three types of perception data. Based on the validity detection results of each perception data, it effectively judges the working status of each sensor, identifies which perception data are reliable, and generates candidate driving paths adapted to the current perception conditions based on the combination of reliable perception data and the determined valid perception data. This ensures that there are still feasible path solutions when some sensors fail. Finally, it optimizes the candidate driving paths based on the local environmental obstacle map and uses the obtained optimal driving path to control the intelligent guided vehicle to drive automatically. This allows the intelligent guided vehicle to continue to obtain valid information and complete navigation control by relying on other normal perception systems when a certain perception system fails, thus realizing reliable autonomous driving of the intelligent guided vehicle in complex environments. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the IGV intelligent autonomous driving method based on multi-mode navigation provided in an embodiment of the present invention.

[0022] Figure 2 This is a schematic diagram of the structure of the IGV intelligent autonomous driving system based on multi-mode navigation provided in an embodiment of the present invention;

[0023] Figure 3 An embodiment diagram of the electronic device provided in this invention;

[0024] Figure 4 An embodiment diagram of a computer-readable storage medium provided in accordance with the present invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0027] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0028] See Figure 1 , Figure 1 This is a flowchart illustrating the IGV intelligent autonomous driving method based on multi-mode navigation provided by the present invention. In this embodiment, the executing entity of the IGV intelligent autonomous driving method based on multi-mode navigation is the autonomous driving control system. The IGV intelligent autonomous driving method based on multi-mode navigation includes:

[0029] Step 10: Based on the lidar sensor, vision sensor, and inertial measurement sensor, collect the first perception data, second perception data, and third perception data of the intelligent guided vehicle, respectively.

[0030] Optionally, during the operation of the intelligent guided vehicle, the autonomous driving control system collects the first perception data, second perception data, and third perception data of the intelligent guided vehicle through lidar sensors, vision sensors, and inertial measurement sensors, respectively.

[0031] Optionally, the lidar sensor acquires three-dimensional point cloud data of the surrounding environment—the first perception data—by emitting a laser beam and receiving the reflected signal. This three-dimensional point cloud data accurately describes the spatial location and shape information of objects in the environment, providing a 3D model of the surrounding environment for the intelligent guided vehicle. For example, in a warehouse environment, the lidar sensor can scan objects such as shelves, goods, and aisles, converting them into three-dimensional point cloud data. Each point in this data contains precise X, Y, and Z coordinate information.

[0032] Optionally, the visual sensor acquires marking line information, i.e., second-level perception data, by capturing images. This marking line information includes the direction and curvature of the marking lines. In factories or warehouses, specific marking lines are typically laid on the floor to guide intelligent guided vehicles. The visual sensor uses image processing technology to identify these marking lines and calculate their direction and curvature. For example, in a factory's transport corridor, the visual sensor captures images of the ground, and through algorithms such as edge detection and color recognition, determines the location and curvature of the marking lines.

[0033] Optionally, the inertial measurement sensor acquires current pose information, i.e., third-party sensing data, by measuring acceleration and angular velocity. Current pose information includes the current position and current heading angle. The inertial measurement sensor can perceive the motion state of the intelligent guided vehicle in real time. For example, during the intelligent guided vehicle's movement, the inertial measurement sensor can continuously measure the vehicle's acceleration and rotation angle to calculate its current position and heading angle.

[0034] Step 20: Perform validity checks on the first, second, and third sensing data to obtain the validity check results for each sensing data.

[0035] Furthermore, the autonomous driving control system performs validity checks on the first perception data, the second perception data, and the third perception data to ensure the accuracy of subsequent navigation decisions.

[0036] For the 3D point cloud data in the first perception data, the autonomous driving control system checks its data integrity and consistency. Regarding data integrity, it checks whether the number of valid point cloud data points in three consecutive scans is less than a preset threshold of 50%. For example, if a LiDAR sensor scans a warehouse shelf area three times consecutively within a certain period, theoretically acquiring 1000 valid point cloud data points per scan, but actually only acquiring 400 each time, which is less than the preset threshold of 50% (i.e., 500), then the 3D point cloud data is deemed to fail in terms of integrity. Regarding consistency, it checks whether the position deviation of the same obstacle obtained from two adjacent scans exceeds 0.5 meters. Assuming the first scan detects the position of a goods in the warehouse as (X1, Y1, Z1), and the second scan changes the position of the same goods to (X2, Y2, Z2), and the calculated position deviation exceeds 0.5 meters, then the 3D point cloud data is deemed to have a consistency problem. If the data integrity or consistency requirements are not met, the validity detection result of the first perception data is deemed invalid; otherwise, the validity detection result is deemed valid.

[0037] For the lane marking information in the second perception data, the autonomous driving control system detects the continuity and accuracy of lane marking recognition. Regarding continuity, it checks whether lane markings cannot be recognized in five consecutive image frames. For example, if the vision sensor captures five consecutive ground images during the intelligent guided vehicle's operation and fails to recognize the preset lane markings, the continuity of lane marking recognition is deemed unsatisfactory. Regarding accuracy, it checks whether the deviation between the recognized lane markings and the preset route exceeds 0.3 meters. If the deviation between the lane markings recognized by the vision sensor and the preset guidance route on the ground reaches 0.4 meters, exceeding the 0.3-meter threshold, the accuracy of lane marking recognition is deemed problematic. If the continuity or accuracy requirements for lane marking recognition are not met, the validity detection result of the second perception data is deemed invalid; otherwise, the validity detection result is deemed valid.

[0038] For the current pose information in the third perception data, the autonomous driving control system detects its drift amount, specifically whether the position drift exceeds 1 meter within 10 seconds. For example, if the drift distance of the vehicle position calculated by the inertial measurement sensor relative to the initial position reaches 1.2 meters within 10 seconds of the intelligent guided vehicle's movement, exceeding the 1-meter threshold, the validity detection result of the third perception data is determined to be invalid; otherwise, the validity detection result is valid.

[0039] Step 30: Based on the validity detection results of each perception data, the valid perception data combination is determined as the input information of the current navigation mode to generate candidate driving paths.

[0040] Furthermore, based on the validity detection results of each perception data point, the autonomous driving control system determines the valid perception data combination and uses it as input information for the current navigation mode to generate candidate driving paths, as described in steps 301 to 304. The valid perception data combination must include at least two valid perception data points; there is no case where only one valid perception data point is included. For example, suppose the first perception data (3D point cloud data) and the second perception data (marker line information) are both valid after validity detection, while the third perception data (current pose information) is invalid. In this case, the autonomous driving control system combines the first and second perception data into a valid perception data combination.

[0041] Step 40: Optimize the candidate driving path based on the local environmental obstacle map constructed from the first perception data to obtain the optimal driving path.

[0042] Furthermore, the autonomous driving control system optimizes the candidate driving path based on the local environmental obstacle map constructed from the first perception data to obtain the optimal driving path, as described in steps 401 to 404. The local environmental obstacle map is constructed based on the three-dimensional point cloud data collected by the lidar sensor, which accurately marks the position and shape of obstacles in the environment.

[0043] Step 50: Control the intelligent guided vehicle to drive automatically based on the optimal driving path.

[0044] Furthermore, the autonomous driving control system translates the optimal driving path into control commands for the intelligent guided vehicle, including speed and steering commands. Based on the path's length and curvature, it calculates the intelligent guided vehicle's speed on different road sections and the steering angle at curves. For example, on straight sections, the autonomous driving control system sends a higher speed command to allow the intelligent guided vehicle to travel quickly; approaching a curve, it reduces speed and sends a corresponding steering command to enable the intelligent guided vehicle to smoothly navigate the curve. Upon receiving these control commands, the intelligent guided vehicle's drive and steering systems execute the corresponding actions, thereby achieving automatic driving along the optimal path and completing the transportation task.

[0045] This invention performs validity checks on three types of sensing data. Based on the validity check results of each sensing data, it effectively judges the working status of each sensor, identifies which sensing data are reliable, and generates candidate driving paths adapted to the current sensing conditions based on the combination of reliable sensing data and the determined valid sensing data. This ensures that there are still feasible path solutions when some sensors fail. Finally, it optimizes the candidate driving paths based on the local environmental obstacle map and uses the obtained optimal driving path to control the intelligent guided vehicle to drive automatically. This allows the intelligent guided vehicle to continue to obtain valid information and complete navigation control by relying on other normal sensing systems when a certain sensing system fails, thus realizing reliable autonomous driving of the intelligent guided vehicle in complex environments.

[0046] In one embodiment, steps 301 to 304 include:

[0047] Step 301: If the first perception data is invalid, but the second and third perception data are valid, then the first valid perception data combination is obtained by combining the marker line information and the current pose information, and a candidate driving path is generated using the first valid perception data combination as input information.

[0048] Optionally, when the validity detection result of the first perception data (3D point cloud data) is invalid, and the validity detection results of the second perception data (marker information) and the third perception data (current pose information) are both valid, the autonomous driving control system combines the valid marker information and the current pose information into the first valid perception data combination. The marker information includes the marker direction and curvature, and the current pose information includes the current position and the current heading angle.

[0049] Furthermore, the autonomous driving control system generates candidate driving paths using the first effective perception data combination as input information, as described in steps 3011 to 3014.

[0050] Step 302: If the second perception data is invalid, and the first and third perception data are valid, then the second valid perception data combination is obtained by combining the three-dimensional point cloud data and the current pose information, and a candidate driving path is generated using the second valid perception data combination as input information.

[0051] Furthermore, when the validity detection result of the second perception data (marker line information) is invalid, and the validity detection results of the first perception data (3D point cloud data) and the third perception data (current pose information) are both valid, the autonomous driving control system combines the valid 3D point cloud data and the current pose information into a second valid perception data combination. The 3D point cloud data provides 3D spatial information of the surrounding environment. Further, the autonomous driving control system uses the second valid perception data combination as input information to generate candidate driving paths, as described in steps 3021 to 3025.

[0052] Step 303: If the third perception data is invalid, and the first and second perception data are valid, then the third valid perception data combination is obtained by combining the three-dimensional point cloud data and the marker line information, and the candidate driving path is generated using the third valid perception data combination as input information.

[0053] Furthermore, when the validity detection result of the third perception data (current pose information) is invalid, and the validity detection results of the first perception data (3D point cloud data) and the second perception data (marker line information) are both valid, the autonomous driving control system combines the valid 3D point cloud data and the marker line information into a third valid perception data combination. Further, the autonomous driving control system uses the third valid perception data combination as input information to generate candidate driving paths, specifically as described in steps 3031 to 3034.

[0054] Step 304: If the first perception data, the second perception data, and the third perception data are valid, then the fourth valid perception data combination is obtained by combining the three-dimensional point cloud data, the marker line information, and the current pose information, and a candidate driving path is generated using the fourth valid perception data combination as input information.

[0055] Furthermore, when the validity detection results of the first perception data (3D point cloud data), the second perception data (marker line information), and the third perception data (current pose information) are all valid, the autonomous driving control system combines the 3D point cloud data, the marker line information, and the current pose information into a fourth valid perception data combination. Further, the autonomous driving control system generates candidate driving paths using the third valid perception data combination as input information, as described in steps 3041 to 3044.

[0056] According to different combinations of effective perception data, the embodiments of the present invention can flexibly generate candidate driving paths suitable for the current environment. This enables the intelligent guided vehicle to plan a candidate driving path that meets environmental constraints and navigation requirements by utilizing effective perception data, regardless of whether the scenario is characterized by only valid marker line information and current pose information, only valid 3D point cloud data and current pose information, only valid 3D point cloud data and marker line information, or all three types of data are valid.

[0057] In one embodiment, steps 3011 to 3014 include:

[0058] Step 3011: Calculate the direction deviation angle based on the direction of the marker line in the marker line information and the current heading angle in the current pose information, and calculate the first tangent direction of the marker line at the current position of the intelligent guided vehicle based on the curvature of the marker line in the marker line information and the current position in the current pose information.

[0059] Optionally, the autonomous driving control system extracts the direction of the road markings from the road marking information and the current heading angle from the current pose information. It then subtracts the angle corresponding to the direction of the road markings from the current heading angle to obtain the direction deviation angle. For example, if the direction of the road markings is due east (corresponding to an angle of 90 degrees), and the current heading angle of the intelligent guided vehicle is 95 degrees, then the direction deviation angle is 95 degrees minus 90 degrees, which is 5 degrees. This indicates that the current driving direction of the intelligent guided vehicle is deviated to the right by 5 degrees relative to the direction of the road markings.

[0060] Furthermore, the autonomous driving control system calculates the first tangent direction of the marker line at the current position of the intelligent guided vehicle based on the marker line curvature in the marker line information and the current position in the current pose information. For a straight marker line (with a curvature of 0), its tangent direction at any position is consistent with the direction of the marker line. For example, for a straight marker line running due east, the first tangent direction at the current position (X0, Y0) is due east (90 degrees). For a curved marker line, the first tangent direction is the tangent angle of the marker line at the current position, which can be determined by the curvature and geometric distribution of the marker line.

[0061] Assume the curvature of a certain curved marker line is 0.02 meters. -1 The current position is located at a specific point on the curve. The tangent direction at this point is calculated to be 85 degrees, so the first tangent direction is 85 degrees.

[0062] Step 3012: Calculate the initial adjustment angle based on the direction deviation angle and the first tangent direction, and extend the initial adjustment angle by a preset distance from the current position to obtain the initial extension point.

[0063] Furthermore, the autonomous driving control system calculates the initial adjustment angle based on the direction deviation angle and the first tangent direction. The initial adjustment angle is equal to the first tangent direction minus the direction deviation angle. This ensures that the intelligent guided vehicle's driving direction aligns with the tangent direction of the marker line at its current position, correcting any directional deviation. For example, if the first tangent direction is 90 degrees and the direction deviation angle is 5 degrees, then the initial adjustment angle is 90 degrees minus 5 degrees, resulting in 85 degrees. This indicates that the intelligent guided vehicle needs to adjust its driving direction to 85 degrees to match the tangent direction of the marker line.

[0064] Furthermore, the autonomous driving control system takes the current position as the starting point and extends it by a preset distance (e.g., 2 meters) according to the initial adjustment angle to obtain the initial extension point. Assuming the current position coordinates are (10 meters, 20 meters), the initial adjustment angle is 90 degrees (due east), and the preset distance is 2 meters, then the coordinates of the initial extension point are (10 meters + 2 meters * cos(90°), 20 meters + 2 meters * sin(90°)), which is (10 meters, 22 meters).

[0065] Step 3013: Calculate the second tangent direction at the initial extension point based on the curvature of the initial extension point and the marker line, and calculate the second adjustment angle based on the initial adjustment angle of the second tangent direction and the initial adjustment angle.

[0066] Furthermore, the autonomous driving control system calculates the second tangent direction at the initial extension point based on the curvature of the marker line. For a straight marker line (with a curvature of 0), the tangent direction at any point is the same; therefore, the second tangent direction is consistent with the first tangent direction. For example, for a straight marker line running due east, the second tangent direction at the initial extension point is still 90 degrees. For a curved marker line, the tangent direction at that point is calculated based on its curvature and the position of the initial extension point on the marker line.

[0067] Assume the curvature of the marker line is 0.02 meters. -1 The initial extension point has moved 2 meters along the marking line relative to the current position, and the second tangent direction at this point is calculated to be 87 degrees.

[0068] Furthermore, the autonomous driving control system calculates a secondary adjustment angle based on the initial adjustment angle and the second tangent direction, where the secondary adjustment angle is equal to the second tangent direction minus the initial adjustment angle. For example, if the second tangent direction is 87 degrees and the initial adjustment angle is 85 degrees, then the secondary adjustment angle is 87 degrees minus 85 degrees, resulting in 2 degrees. This indicates that from the initial extension point, the driving direction needs to be adjusted by another 2 degrees to match the tangent direction of the marker line at that point.

[0069] Step 3014: Starting from the initial extension point, extend the path by a preset distance according to the second adjustment angle to obtain the second extension point. Connect the current position, the initial extension point, and the second extension point to obtain the candidate driving path.

[0070] Furthermore, the autonomous driving control takes the initial extension point as the starting point and extends it by a preset distance (the same preset distance as in step 3012, such as 2 meters) according to the second adjustment angle to obtain the second extension point. Assuming the coordinates of the initial extension point are (10 meters, 22 meters), the second adjustment angle is 90 degrees (due east), and the preset distance is 2 meters, then the coordinates of the second extension point are (10 meters + 2 meters * cos(90°), 22 meters + 2 meters * sin(90°)), i.e., (10 meters, 24 meters). Further, the autonomous driving control connects the current position, the initial extension point, and the second extension point sequentially to form a continuous broken line. This broken line is the candidate driving path generated based on the first effective perception data combination. For example, connecting the current position (10 meters, 20 meters), the initial extension point (10 meters, 22 meters), and the second extension point (10 meters, 24 meters) forms a straight path along the due east direction, which serves as the candidate driving path.

[0071] Based on valid marker line information and current pose information, this invention extends and connects the path in stages by calculating directional deviation, tangent direction, and adjusting angle to form candidate driving paths. This ensures that the generated candidate driving paths closely follow the direction and curvature characteristics of the marker lines, while correcting the deviation between the current heading angle of the intelligent guided vehicle and the direction of the marker lines. This ensures that the path conforms to the guiding trend of the marker lines. Furthermore, by extending and adjusting the angle twice, the generated candidate driving paths can better adapt to the straight or curved shape of the marker lines, providing the intelligent guided vehicle with a basic driving path that meets the navigation direction requirements and improving the accurate autonomous driving of the intelligent guided vehicle in complex environments.

[0072] In one embodiment, steps 3021 to 3025 include:

[0073] Step 3021: Determine the obstacle outline boundary points of all obstacles based on the local environment obstacle map, and calculate the boundary distance from the current position of the intelligent guided vehicle to each obstacle outline boundary point based on the boundary coordinates of each obstacle outline boundary point and the position coordinates of the current position in the current position information.

[0074] Optionally, the autonomous driving control system constructs a local environment obstacle map based on 3D point cloud data and extracts the contour boundary points of all obstacles. Each obstacle contour boundary point contains its boundary coordinates (Xi, Yi, Zi) in the environmental coordinate system, where i is the index of the obstacle contour boundary point. For example, in a warehouse environment, a shelf acts as an obstacle, and its contour boundary points may include four bottom corner points: (5m, 10m, 0m), (7m, 10m, 0m), (7m, 15m, 0m), and (5m, 15m, 0m). Simultaneously, the autonomous driving control system obtains the current position coordinates (X0, Y0, Z0) from the current pose information. For each obstacle contour boundary point, the autonomous driving control system calculates the boundary distance from the current position of the intelligent guided vehicle to that point using the distance calculation formula between two points.

[0075] For example, if the current location coordinates are (3m, 12m, 0m) and the coordinates of a certain obstacle boundary point are (5m, 10m, 0m), then the boundary distance of that point is D≈2.83m.

[0076] Step 3022: Determine the target obstacle contour boundary point with the smallest boundary distance based on the boundary distance of each obstacle contour boundary point, and calculate the intersection point of the target obstacle contour boundary point with the direction of extension of the current heading angle based on the target obstacle contour boundary point and the current heading angle in the current position information.

[0077] Furthermore, the autonomous driving control system compares all boundary distances D and selects the obstacle contour boundary point with the smallest boundary distance, thus identifying it as the target obstacle contour boundary point. For example, after comparison, the boundary distance of a certain obstacle contour boundary point (5 meters, 12 meters, 0 meters) is the smallest, at 2 meters, therefore this point is identified as the target obstacle contour boundary point.

[0078] Furthermore, the autonomous driving control system uses the current position as the starting point and draws an extended straight line L along the current heading angle θ0 (e.g., 90 degrees, i.e., due east). It calculates the intersection point P of this straight line L with the outline boundary of the obstacle where the target obstacle's outline boundary point is located. If the target obstacle is a rectangular shelf with a boundary of x = 5 meters (from y = 10 meters to y = 15 meters), the current heading angle is 90 degrees (due east), and the current position is (3 meters, 12 meters, 0 meters), then the equation of the extended straight line L is y = 12 meters (due east is the direction of increasing x), and the coordinates of the intersection point P of this line and the shelf boundary x = 5 meters are (5 meters, 12 meters, 0 meters).

[0079] Step 3023: Calculate the direction angle of the line connecting the intersection point and the current position based on the intersection point coordinates and the position coordinates, and calculate the turning angle based on the current heading angle and the direction angle of the line.

[0080] Furthermore, the autonomous driving control system calculates the direction angle θp of the line connecting the intersection point P and the current position coordinates (Xp, Yp, Zp), as detailed in the angle calculation formula, which will not be elaborated here. Additionally, the autonomous driving control system performs angle corrections based on the coordinate quadrants to ensure the angle remains within the range of 0-360 degrees.

[0081] For example, if the coordinates of the intersection point P are (5 meters, 12 meters, 0 meters) and the current position coordinates are (3 meters, 12 meters, 0 meters), then the direction angle of the connecting line θp = 0 degrees (or 360 degrees), which is due east.

[0082] Furthermore, the autonomous driving control system calculates the steering angle based on the current heading angle and the direction angle of the connecting line. The steering angle is equal to the direction angle of the connecting line minus the current heading angle, i.e., Δθ = θp - θ0. If the calculation result is positive, it indicates a right turn; if it is negative, it indicates a left turn.

[0083] For example, if the current heading angle θ0 is 90 degrees (due east) and the direction angle of the connecting line θp is 90 degrees, then the turning angle Δθ = 90 degrees - 90 degrees = 0 degrees; if the direction angle of the connecting line θp is 120 degrees, then the turning angle Δθ = 120 degrees - 90 degrees = 30 degrees, that is, it is necessary to turn 30 degrees to the right.

[0084] Step 3024: Starting from the current position, extend the distance by the turning angle to obtain the obstacle avoidance starting point. Starting from the obstacle avoidance starting point, traverse the local environment obstacle map along the direction of avoiding obstacles, and select target points in the obstacle-free area in turn.

[0085] Furthermore, the autonomous driving control system uses the current position as the starting point and extends it by a preset distance (e.g., 1.5 meters) according to the steering angle to obtain the obstacle avoidance starting point A. Assuming the current position coordinates are (3 meters, 12 meters, 0 meters), the steering angle is 30 degrees, and the preset distance is 1.5 meters, the coordinates of the obstacle avoidance starting point A are calculated as follows:

[0086] Xa=X0+1.5*cos(30°)≈3+1.5*0.866≈4.3 meters.

[0087] Ya=Y0+1.5*sin(30°)≈12+1.5*0.5≈12.75 meters.

[0088] Subsequently, the autonomous driving control system, starting from obstacle avoidance point A, traverses the local obstacle map along the direction of obstacle avoidance (the direction consistent with the steering angle). During the traversal, within the obstacle-free safe area, the system sequentially selects target points B, C, D, etc., at preset intervals (e.g., 2 meters). For example, starting from obstacle avoidance point A (4.3 meters, 12.75 meters, 0 meters), it continues to extend along a 30-degree direction, selecting target point B in the obstacle-free area 2 meters away from point A, then selecting target point C in the obstacle-free area 2 meters away from point B, and so on.

[0089] Step 3025: Connect the current position, obstacle avoidance starting point, and target point in sequence to obtain candidate driving paths.

[0090] Furthermore, the autonomous driving control system sequentially connects the current position, obstacle avoidance starting point, and all selected target points to form a continuous broken line. This broken line is the candidate driving path generated based on the combination of the second effective perception data. For example, connecting the current position (3m, 12m, 0m), obstacle avoidance starting point A (4.3m, 12.75m, 0m), target point B (6.0m, 13.5m, 0m), and target point C (7.7m, 14.25m, 0m) sequentially forms a broken line path that turns right to avoid obstacles, serving as a candidate driving path.

[0091] Based on effective 3D point cloud data and current pose information, this invention identifies the nearest obstacle, calculates the obstacle avoidance steering angle, plans the obstacle avoidance path, and selects a target point to generate a candidate driving path that avoids obstacles. This candidate driving path starts from the current position, avoids the nearest obstacle by steering, and extends into an obstacle-free area, ensuring path safety. Simultaneously, the path planning process incorporates the current heading angle information, making the steering operation conform to the vehicle's driving characteristics. The generated candidate driving path can effectively avoid obstacles in the local environment, providing a safe and feasible driving route selection for intelligent guided vehicles, meeting navigation needs in situations without marked lines, and improving the accurate autonomous driving of intelligent guided vehicles in complex environments.

[0092] In one embodiment, steps 3031 to 3034 include:

[0093] Step 3031: Determine the bounding rectangles of all obstacles based on the local environment obstacle map, and calculate the potential intersection area between each marker line and the bounding rectangle of the obstacle based on the vertex coordinates of the bounding rectangles and the direction of the marker lines in the marker line information.

[0094] Optionally, the autonomous driving control system constructs a local environment obstacle map based on 3D point cloud data, generating a bounding rectangle for each obstacle. The bounding rectangle is the smallest rectangle that can completely enclose the obstacle, and its vertex coordinates are calculated from the boundary points of the obstacle's outline.

[0095] For example, the outline boundary point coordinates of a shelf obstacle are (5m, 10m, 0m), (7m, 10m, 0m), (7m, 15m, 0m), (5m, 15m, 0m), and its circumscribed rectangle coincides with the obstacle itself, with vertex coordinates of (5m, 10m, 0m), (7m, 10m, 0m), (7m, 15m, 0m), (5m, 15m, 0m).

[0096] Furthermore, the autonomous driving control system extracts the direction of the road markings (e.g., due east, 90 degrees) from the marking information and determines their spatial distribution in the environment based on the geometric parameters of the markings (start point, end point, curvature, etc.). Further, the autonomous driving control system determines potential intersection areas by calculating the intersection of the extended trajectory of the markings with the circumscribed rectangle of each obstacle. A potential intersection area refers to the region where the markings, if extended along their original direction, would intersect with the circumscribed rectangle of an obstacle. For example, an eastward-oriented marking line extending from (3m, 12m, 0m) to (8m, 12m, 0m) intersects with the aforementioned shelf circumscribed rectangles from (5m, 10m, 0m) to (7m, 15m, 0m) at a distance of (5m, 12m, 0m) to (7m, 12m, 0m); this region is a potential intersection area.

[0097] Step 3032: Based on the curvature of the signage line in the potential intersection area and the signage line information, calculate the bending offset of the signage line in each potential intersection area, and determine the unobstructed offset direction of the signage line based on the bending offset and the potential intersection area.

[0098] Furthermore, the autonomous driving control system calculates the bending offset of the lane markings within each potential intersection area based on the lane marking curvature information. The bending offset refers to the positional shift within the intersection area caused by the lane markings' own curvature, where the bending offset δ = r * (1 - cosθ), where r is the radius of curvature of the lane marking (the reciprocal of curvature), and θ is the bending angle of the lane marking within the intersection area. For example, the lane marking curvature is 0.02 meters. -1 (Radius of curvature r = 50 meters), the bending angle θ = 5 degrees in the potential intersection area, then the bending offset δ = 50 * (1 - cos5°) ≈ 0.2 meters.

[0099] Furthermore, the autonomous driving control system analyzes whether the marker line will avoid obstacles when it shifts to the left or right based on the bending offset and the spatial location of the potential intersection area. The direction of the obstacle-free offset is determined by comparing the unobstructed space on both sides. For example, for the shelf obstacles corresponding to the potential intersection areas (5m, 12m, 0m) to (7m, 12m, 0m), if the space on the left (y<12m) is narrow and the space on the right (y>12m) is open, then the direction of the obstacle-free offset is determined to be to the right.

[0100] Step 3033: Based on the unobstructed offset direction and the direction of the marking line, calculate the initial direction angle of the offset path, and based on the initial direction angle and the potential intersection area, calculate the safe distance boundary point between the offset path and the bounding rectangle of the obstacle.

[0101] Furthermore, the autonomous driving control system calculates the initial direction angle of the offset path based on the obstacle-free offset direction and the direction of the marking lines. The initial direction angle is equal to the angle correction value corresponding to the offset direction of the marking lines. For example, if the marking lines are 90 degrees (due east), the obstacle-free offset direction is to the right (clockwise), and the correction angle is 10 degrees, then the initial direction angle is 90 degrees + 10 degrees = 100 degrees.

[0102] Furthermore, the autonomous driving control system calculates the safe distance boundary point between the offset path and the circumscribed rectangle of the obstacle based on the initial direction angle and potential intersection areas. The safe distance boundary point refers to the point on the offset path that maintains a preset safe distance (e.g., 0.8 meters) from the circumscribed rectangle of the obstacle. For example, for the aforementioned shelf circumscribed rectangle (5m, 10m, 0m) to (7m, 15m, 0m), the preset safe distance is 0.8 meters. The point on the offset path extending along the initial direction angle of 100 degrees that maintains a safe distance of 0.8 meters from the right side of the shelf (x = 7m) is the safe distance boundary point, and its coordinates are calculated as (7.8m * cos(10 degrees), 12m + 7.8m * sin(10 degrees)) ≈ (7.7m, 13.36m, 0m).

[0103] Step 3034: Based on the safety distance boundary points and the curvature of the marking lines, calculate the curve correction parameters of the offset path, and smoothly transition the marking lines according to the initial direction angle offset combined with the curve correction parameters. Connect each safety distance boundary point in sequence to form a continuous curve to obtain the candidate driving path.

[0104] Furthermore, the autonomous driving control system calculates curve correction parameters for the off-track path based on the safety distance boundary points and the curvature of the lane markings. These curve correction parameters ensure a smooth transition back to the original lane markings after avoiding obstacles; their values ​​are determined based on the lane marking curvature and the position of the safety distance boundary points. For example, the lane marking curvature is 0.02 meters. -1 If the offset of the safety distance boundary point from the original marking line is 1 meter, then the curve correction parameter is 0.015 meters. -1 This makes the curvature of the offset path slightly less than that of the original marking line.

[0105] Furthermore, the autonomous driving control system offsets the lane markings according to the initial direction angle and uses curve correction parameters to smoothly transition the path, ensuring continuous curvature changes as the path enters and leaves potential intersections. Further, the autonomous driving control system sequentially connects each safety distance boundary point to form a continuous curve, which is the candidate driving path generated based on the combination of third-party effective perception data. For example, connecting the points on the current lane markings, the safety distance boundary points, and the points transitioning back to the original lane markings sequentially forms a continuous curved path that shifts to the right to avoid shelf obstacles and then smoothly returns to the original direction.

[0106] Based on effective 3D point cloud data and marker line information, this invention identifies the intersection areas of marker lines and obstacles, determines the obstacle avoidance offset direction, calculates offset path parameters, and performs smooth transition processing. Ultimately, it generates candidate driving paths that both follow marker line guidance and avoid obstacles. This ensures that the generated candidate driving paths extend along the marker lines in obstacle-free areas and avoid obstacles through reasonable offsets when approaching them, while maintaining the smoothness and continuity of the path. It balances the accuracy of navigation direction with driving safety, providing a reliable driving path selection for intelligent guided vehicles in the absence of precise pose information. This invention also demonstrates the reliable autonomous driving capabilities of intelligent guided vehicles in complex environments.

[0107] In one embodiment, steps 3041 to 3044 include:

[0108] Step 3041: Based on the location coordinates of the current location in the current location information and the direction of the marker line in the marker line information, calculate the lateral distance between the intelligent guided vehicle and the marker line at the current location, and based on the lateral distance and the local environmental obstacle map, determine the target obstacle located between the current location of the intelligent guided vehicle and the marker line.

[0109] Optionally, the autonomous driving control system obtains the current position coordinates (X0, Y0, Z0) from the current pose information and extracts the direction (e.g., due east, 90 degrees) and geometric parameters of the marker line information. Lateral distance refers to the vertical distance from the current position of the intelligent guided vehicle to the marker line. The calculation formula is: if the equation of the marker line is ax + by + c = 0, then the lateral distance is calculated according to the formula for the distance from a point to a line, which will not be elaborated here. For example, if the marker line is a straight line due east along y = 12 meters (equation y - 12 = 0), and the current position coordinates are (5 meters, 14 meters, 0 meters), then the lateral distance d2 meters indicates that the intelligent guided vehicle is located 2 meters north of the marker line.

[0110] Furthermore, the autonomous driving control system, based on a local obstacle map, retrieves all obstacles located between the current position and the marker line (i.e., obstacles distributed along the lateral distance direction) and identifies them as target obstacles. For example, if the 3D point cloud data shows a shelf obstacle between the current position (5m, 14m, 0m) and the marker line y = 12m, with an outline range of (4m, 12.5m, 0m) to (6m, 13.5m, 0m), then this shelf is identified as a target obstacle.

[0111] Step 3042: Based on the current heading angle in the target obstacle and current position information, calculate the conflict angle between each target obstacle and the current heading of the intelligent guided vehicle, and determine the width threshold of the safe passage based on the conflict angle and the curvature of the marking line in the marking line information.

[0112] Furthermore, the autonomous driving control system calculates the conflict angle α between each target obstacle and the current heading angle θ0 (e.g., 90 degrees, due east) based on the outline boundary of the target obstacle and the current heading angle. The conflict angle refers to the projection angle of the target obstacle in the current heading direction. The formula for calculating the conflict angle α is: α=arctan[(Ymax-Ymin) / (Xmax-Xmin)], where (Xmin,Ymin) and (Xmax,Ymax) are the coordinates of the diagonal vertices of the bounding rectangle of the target obstacle.

[0113] For example, if the circumscribed rectangle vertices of the above-mentioned shelf obstacle are (4m, 12.5m, 0m) and (6m, 13.5m, 0m), then the conflict angle α≈26.56 degrees.

[0114] Furthermore, the autonomous driving control system combines the collision angle α and the curvature k of the marker line (e.g., 0.01 meters). -1 To determine the width threshold W of the safe passageway, the width threshold must be such that it avoids target obstacles while also accommodating the curvature of the marking lines. The formula for calculating the width threshold W is: W = 2 * (D + r * sinα), where D is 1.5 times the width of the intelligent guided vehicle (e.g., if the intelligent guided vehicle is 1.2 meters wide, then D = 1.8 meters), and r is the radius of curvature of the marking line (r = 1 / k).

[0115] For example, if D = 1.8 meters, r = 100 meters, and α = 26.56 degrees, then W = 2*(1.8 + 100*sin26.56°) ≈ 93 meters, ensuring that the passage is wide enough to accommodate vehicles and avoid obstacles.

[0116] Step 3043: Based on the width threshold and lateral distance, calculate the feasible offset for the intelligent guided vehicle to move closer to the marking line, and determine the target position after the offset based on the feasible offset and the direction of the marking line.

[0117] Furthermore, the autonomous driving control system calculates the feasible offset Δd based on the width threshold W and the lateral distance d. The feasible offset refers to the maximum safe distance that the intelligent guided vehicle can approach the marked line, which must satisfy Δd≤d and Δd≤W / 2 (ensuring that it is within the safe passage).

[0118] For example, if the lateral distance d = 2 meters and the width threshold W = 5 meters (W / 2 = 2.5 meters), then the feasible offset Δd = 2 meters (i.e., completely close to the marking line). If the lateral distance d = 3 meters and the width threshold W = 5 meters (W / 2 = 2.5 meters), then the feasible offset Δd = 2.5 meters (limited by the width of the safety passage).

[0119] Furthermore, the autonomous driving control system determines the target position (Xt, Yt, Zt) after the offset based on the feasible offset Δd and the direction of the marker line. The target position is the position after moving Δd along the lateral distance direction towards the marker line from the current position. The formula for calculating the target position is: if the direction of the marker line is θ and the lateral direction is θ+90 degrees, then Xt=X0-Δd*sin(θ+90°), Yt=Y0+Δd*cos(θ+90°).

[0120] For example, if the direction of the marker line is θ = 90 degrees (due east) and the lateral direction is 180 degrees (due south), and the current position is (5 meters, 14 meters), with a feasible offset Δd = 2 meters, then the target position Xt = 5 - 2*sin(180°) = 5 meters and Yt = 14 + 2*cos(180°) = 12 meters, which coincides with the marker line.

[0121] Step 3044: Based on the target location coordinates and the curvature of the marker line, calculate the transition curve from the current location to the target location. Starting from the current location, travel along the transition curve to the target location, and then continue to extend along the marker line direction and curvature to form a continuous path, thus obtaining a candidate travel path.

[0122] Furthermore, the autonomous driving control system calculates a transition curve from the current position (X0, Y0, Z0) to the target position based on the target position coordinates (Xt, Yt, Zt) and the curvature k of the marker line. The transition curve employs a smooth curve that matches the curvature of the marker line, and its parametric equation is as follows: Where s is the curve length parameter, and L is the total length of the transition curve (e.g., 5 meters). For example, the current position is (5 meters, 14 meters, 0 meters), the target position is (5 meters, 12 meters, 0 meters), the curvature of the marker line is k = 0, and the transition curve is a straight line extending 5 meters south from (5 meters, 14 meters, 0 meters) to (5 meters, 12 meters, 0 meters).

[0123] Furthermore, starting from the current position, the autonomous driving control system travels along the transition curve to the target position, and then continues to extend the path according to the direction and curvature of the marking lines, ensuring a smooth connection between the path and the marking lines at the target position. For example, after reaching the target position (5 meters, 12 meters, 0 meters), it extends along the marking lines in a due east direction (90 degrees) with a curvature of 0, forming a continuous path. Ultimately, this continuous path composed of the transition curve and the extended segment of the marking lines is the candidate driving path.

[0124] This invention determines the relative position to the marking line by calculating the lateral distance, identifies intermediate obstacles and calculates a safe passage, plans a reasonable offset to bring the vehicle closer to the marking line, and finally connects them through a smooth transition curve to form a continuous path. This ensures that the generated candidate driving path can follow the marking line guidance to the maximum extent, effectively avoid obstacles, and maintain coordination with the vehicle's current position and heading. This provides the intelligent guided vehicle with a safe, accurate driving path selection that conforms to the navigation intent, improving the reliable and accurate autonomous driving of the intelligent guided vehicle in complex environments.

[0125] In one embodiment, steps 401 to 404 include:

[0126] Step 401: Discretize the candidate driving path to obtain the initial path points.

[0127] Optionally, the autonomous driving control system discretizes the candidate driving path, taking points along the path at preset intervals (e.g., 0.5 meters) to obtain a series of initial path points. Each initial path point contains its coordinate information (Xj, Yj, Zj) in the environmental coordinate system, where j is the sequence number of the initial path point. For example, a straight path extending eastward from (10m, 20m, 0m) to (10m, 30m, 0m), after discretization at 0.5-meter intervals, yields 21 initial path points: (10m, 20m, 0m), (10m, 20.5m, 0m), (10m, 21m, 0m)...(10m, 30m, 0m). For curved candidate driving paths, points are also taken along the curve at preset intervals to form a sequence of initial path points.

[0128] Step 402: Perform collision detection on the initial path points based on the local environment obstacle map to determine whether there are target path points within the obstacle grid of the local environment obstacle map among the initial path points.

[0129] Furthermore, the autonomous driving control system divides the local obstacle map into a grid structure (e.g., a 0.1m x 0.1m grid) and marks the obstacle grids containing obstacles. For each initial path point (Xj, Yj, Zj), the autonomous driving control system determines whether the grid containing that point is an obstacle grid. If the coordinates of the initial path point fall within the range of an obstacle grid, it is determined that there is a collision risk between that point and an obstacle. For example, in the local obstacle map, the range of an obstacle grid is (10m ≤ X ≤ 10.5m, 25m ≤ Y ≤ 25.5m). If the coordinates of the initial path point (10m, 25.3m, 0m) fall within this grid, it is determined that the initial path point is within the obstacle grid and there is a collision risk, thus identifying it as the target path point.

[0130] Step 403: If a target path point exists, remove the target path point from the initial path points to obtain the updated path point.

[0131] Furthermore, when a target pathpoint is detected, the autonomous driving control system removes all identified target pathpoints from the initial pathpoint set, and the remaining initial pathpoints form the updated pathpoint set. For example, if the initial pathpoint set contains 21 points, and three points (10m, 25m, 0m), (10m, 25.5m, 0m), and (10m, 26m, 0m) are identified as target pathpoints, then after removing them, 18 updated pathpoints are obtained.

[0132] If no target path point is detected (i.e., all initial path points are not within the obstacle grid), the updated path point is exactly the same as the initial path point, and no culling operation is required.

[0133] Step 404: With the goal of minimizing the total path distance, optimize the path by combining the path distances between path points after the update, and obtain the optimal driving path.

[0134] Furthermore, the autonomous driving control system uses minimizing the total path distance as its optimization objective and reconstructs the path based on the updated path points, specifically as follows:

[0135] The autonomous driving control system calculates the path distance between adjacent points in the updated path points. The formula for calculating the path distance can be obtained using the formula for calculating the distance between two points. Furthermore, the autonomous driving control system adjusts the connection method between path points (ensuring that connecting lines do not pass through obstacle grids) to minimize the total path distance formed by connecting all updated path points. For example, if the updated path points are A (10m, 20m, 0m), B (10m, 24m, 0m), C (10m, 27m, 0m), and D (10m, 30m, 0m), the straight-line distances between each point are calculated and connected sequentially to form a continuous path with the minimum total path distance. If there is an obstacle on the straight line between adjacent path points, a new transition point is inserted between the two points to ensure that the path bypasses the obstacle and minimizes the total distance. Finally, the continuous path with the minimum total path distance and no collision risk obtained after optimization is the optimal driving path.

[0136] This invention, through discretizing path points, detecting and eliminating collision points, and reconstructing the path with the goal of minimizing the total distance, ensures that the generated optimal driving path can completely avoid obstacles in the environment while having the shortest driving distance. Therefore, the optimized path takes into account both driving safety and efficiency, providing the optimal driving route for the intelligent guided vehicle and ensuring that the intelligent guided vehicle can safely reach the target location.

[0137] Furthermore, the IGV intelligent autonomous driving system based on multi-mode navigation provided by the present invention will be described below. The IGV intelligent autonomous driving system based on multi-mode navigation described below can be referred to in correspondence with the IGV intelligent autonomous driving method based on multi-mode navigation described above.

[0138] Optional, refer to Figure 2 , Figure 2 This is a schematic diagram of the structure of the IGV intelligent autonomous driving system based on multi-mode navigation provided by the present invention. The IGV intelligent autonomous driving system based on multi-mode navigation includes...

[0139] The data acquisition module 210 is used to collect the first perception data, the second perception data, and the third perception data of the intelligent guided vehicle based on the lidar sensor, the vision sensor, and the inertial measurement sensor, respectively.

[0140] The data detection module 220 is used to perform validity detection on the first sensing data, the second sensing data and the third sensing data, and obtain the validity detection results of each sensing data.

[0141] The driving path generation module 230 is used to combine the valid perception data determined by the validity detection results of each perception data into the input information of the current navigation mode to generate candidate driving paths;

[0142] The driving path optimization module 240 is used to optimize candidate driving paths based on the local environment obstacle map constructed from the first perception data to obtain the optimal driving path.

[0143] The driving control module 250 is used to control the automatic driving of the intelligent guided vehicle based on the optimal driving path.

[0144] This invention performs validity checks on three types of sensing data. Based on the validity check results of each sensing data, it effectively judges the working status of each sensor, identifies which sensing data are reliable, and generates candidate driving paths adapted to the current sensing conditions based on the combination of reliable sensing data and the determined valid sensing data. This ensures that there are still feasible path solutions when some sensors fail. Finally, it optimizes the candidate driving paths based on the local environmental obstacle map and uses the obtained optimal driving path to control the intelligent guided vehicle to drive automatically. This allows the intelligent guided vehicle to continue to obtain valid information and complete navigation control by relying on other normal sensing systems when a certain sensing system fails, thus realizing reliable autonomous driving of the intelligent guided vehicle in complex environments.

[0145] Please see Figure 3 , Figure 3 An embodiment diagram of an electronic device provided in accordance with the present invention. For example... Figure 3As shown, this embodiment of the invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, it performs the following steps:

[0146] Based on lidar sensors, vision sensors, and inertial measurement sensors, the first perception data, second perception data, and third perception data of the intelligent guided vehicle are collected respectively.

[0147] The validity of the first, second, and third sensing data is checked to obtain the validity check results for each sensing data.

[0148] The valid perception data determined based on the validity detection results of each perception data is used as the input information for the current navigation mode to generate candidate driving paths;

[0149] The local environmental obstacle map constructed based on the first perception data is used to optimize the candidate driving path to obtain the optimal driving path;

[0150] The intelligent guided vehicle drives automatically based on the optimal driving path control.

[0151] Please see Figure 4 , Figure 4 An embodiment diagram of a computer-readable storage medium provided in accordance with an embodiment of the present invention is shown. Figure 4 As shown, this embodiment provides a computer-readable storage medium 400 on which a computer program 311 is stored. When the computer program 311 is executed by a processor, it performs the following steps:

[0152] Based on lidar sensors, vision sensors, and inertial measurement sensors, the first perception data, second perception data, and third perception data of the intelligent guided vehicle are collected respectively.

[0153] The validity of the first, second, and third sensing data is checked to obtain the validity check results for each sensing data.

[0154] The valid perception data determined based on the validity detection results of each perception data is used as the input information for the current navigation mode to generate candidate driving paths;

[0155] The local environmental obstacle map constructed based on the first perception data is used to optimize the candidate driving path to obtain the optimal driving path;

[0156] The intelligent guided vehicle drives automatically based on the optimal driving path control.

[0157] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the IGV intelligent autonomous driving method based on multi-mode navigation provided by the above methods, the method comprising:

[0158] Based on lidar sensors, vision sensors, and inertial measurement sensors, the first perception data, second perception data, and third perception data of the intelligent guided vehicle are collected respectively.

[0159] The validity of the first, second, and third sensing data is checked to obtain the validity check results for each sensing data.

[0160] The valid perception data determined based on the validity detection results of each perception data is used as the input information for the current navigation mode to generate candidate driving paths;

[0161] The local environmental obstacle map constructed based on the first perception data is used to optimize the candidate driving path to obtain the optimal driving path;

[0162] The intelligent guided vehicle drives automatically based on the optimal driving path control.

[0163] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0164] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0165] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An IGV intelligent autonomous driving method based on multi-mode navigation, characterized in that, include: Based on lidar sensors, vision sensors, and inertial measurement sensors, the first perception data, second perception data, and third perception data of the intelligent guided vehicle are collected respectively. The validity of the first sensing data, the second sensing data, and the third sensing data is checked to obtain the validity check results of each sensing data; the first sensing data includes three-dimensional point cloud data; the second sensing data includes marker line information; and the third sensing data includes current pose information. The valid perception data determined based on the validity detection results of each perception data is used as the input information for the current navigation mode to generate candidate driving paths; The candidate driving path is optimized based on the local environmental obstacle map constructed from the first perception data to obtain the optimal driving path. The intelligent guided vehicle is controlled to drive automatically based on the optimal driving path. If the first perception data is invalid, but the second and third perception data are valid, then the first valid perception data combination is obtained by combining the identification line information and the current pose information, and the candidate driving path is generated using the first valid perception data combination as input information. Generating the candidate driving path using the first valid perception data combination as input information includes: Based on the direction of the marker line in the marker line information and the current heading angle in the current pose information, the direction deviation angle is calculated, and based on the curvature of the marker line in the marker line information and the current position in the current pose information, the first tangential direction of the marker line at the current position of the intelligent guided vehicle is calculated. Calculate the initial adjustment angle based on the directional deviation angle and the first tangent direction, and extend a preset distance from the current position according to the initial adjustment angle to obtain the initial extension point; The second tangent direction at the initial extension point is calculated based on the initial extension point and the curvature of the marker line, and the second adjustment angle is calculated based on the second tangent direction and the initial adjustment angle. Starting from the initial extension point, extend the path by a preset distance according to the second adjustment angle to obtain a second extension point. Connect the current position, the initial extension point, and the second extension point to obtain the candidate driving path.

2. The IGV intelligent autonomous driving method based on multi-mode navigation according to claim 1, characterized in that, The valid perception data determined based on the validity detection results of each perception data is combined into the input information of the current navigation mode to generate candidate driving paths, including: If the second perception data is invalid, and the first and third perception data are valid, then the third point cloud data and the current pose information are combined to obtain the second valid perception data combination, and the candidate driving path is generated using the second valid perception data combination as input information. If the third perception data is invalid, and the first and second perception data are valid, then the third valid perception data combination is obtained by combining the three-dimensional point cloud data and the marking line information, and the candidate driving path is generated by using the third valid perception data combination as input information. If the first, second, and third perception data are valid, then the fourth valid perception data combination is obtained by combining the three-dimensional point cloud data, the marker line information, and the current pose information, and the candidate driving path is generated using the fourth valid perception data combination as input information.

3. The IGV intelligent autonomous driving method based on multi-mode navigation according to claim 2, characterized in that, The step of generating the candidate driving path using the second effective perception data combination as input information includes: Based on the local environment obstacle map, the obstacle outline boundary points of all obstacles are determined, and based on the boundary coordinates of each obstacle outline boundary point and the position coordinates of the current position in the current position information, the boundary distance from the current position of the intelligent guided vehicle to each obstacle outline boundary point is calculated. Based on the boundary distance of each obstacle contour boundary point, determine the target obstacle contour boundary point with the smallest boundary distance, and based on the target obstacle contour boundary point and the current heading angle in the current position information, calculate the intersection point of the target obstacle contour boundary point with the direction of extension of the current heading angle. Calculate the direction angle of the line connecting the intersection point and the current position based on the intersection point coordinates and the position coordinates, and calculate the turning angle based on the current heading angle and the direction angle of the line. Starting from the current position, extend a preset distance according to the turning angle to obtain the obstacle avoidance starting point. Starting from the obstacle avoidance starting point, traverse the local environment obstacle map along the direction of avoiding obstacles, and select target points in the obstacle-free area in sequence. The candidate driving path is obtained by sequentially connecting the current position, the obstacle avoidance starting point, and the target point.

4. The IGV intelligent autonomous driving method based on multi-mode navigation according to claim 2, characterized in that, The step of generating the candidate driving path using the third effective perception data combination as input information includes: Based on the local environment obstacle map, the bounding rectangles of all obstacles are determined, and based on the vertex coordinates of the bounding rectangles of the obstacles and the direction of the marker lines in the marker line information, the potential intersection area between each marker line and the bounding rectangle of the obstacle is calculated. Based on the potential intersection area and the curvature of the marker line in the marker line information, the bending offset of the marker line in each potential intersection area is calculated, and based on the bending offset and the potential intersection area, the unobstructed offset direction of the marker line is determined. Based on the unobstructed offset direction and the direction of the marking line, the initial direction angle of the offset path is calculated, and based on the initial direction angle and the potential intersection area, the safe distance boundary point between the offset path and the bounding rectangle of the obstacle is calculated. Based on the safety distance boundary points and the curvature of the marker lines, the curve correction parameters of the offset path are calculated, and the marker lines are smoothly transitioned according to the initial direction angle offset combined with the curve correction parameters. The safety distance boundary points are then connected sequentially to form a continuous curve, thus obtaining the candidate driving path.

5. The IGV intelligent autonomous driving method based on multi-mode navigation according to claim 2, characterized in that, The step of generating the candidate driving path using the fourth effective perception data combination as input information includes: Based on the location coordinates of the current location in the current location information and the direction of the marker line in the marker line information, the lateral distance between the intelligent guided vehicle and the marker line at the current location is calculated, and based on the lateral distance and the local environmental obstacle map, the target obstacle located between the current location of the intelligent guided vehicle and the marker line is determined. Based on the target obstacle and the current heading angle in the current position information, calculate the conflict angle between each target obstacle and the current heading of the intelligent guided vehicle, and determine the width threshold of the safe passage based on the conflict angle and the curvature of the marking line in the marking line information. Based on the width threshold and the lateral distance, calculate the feasible offset for the intelligent guided vehicle to move closer to the marking line, and determine the target position after the offset based on the feasible offset and the direction of the marking line. Based on the target location coordinates and the curvature of the marker line, a transition curve from the current location to the target location is calculated. Starting from the current location, the vehicle travels along the transition curve to the target location and continues to extend along the marker line direction and the marker line curvature to form a continuous path, thus obtaining the candidate driving path.

6. The IGV intelligent autonomous driving method based on multi-mode navigation according to any one of claims 1 to 5, characterized in that, The local environmental obstacle map constructed based on the first perception data is used to optimize the candidate driving path to obtain the optimal driving path, including: The candidate driving paths are discretized to obtain initial path points; Collision detection is performed on the initial path points based on the local environment obstacle map to determine whether there are target path points within the obstacle grid of the local environment obstacle map among the initial path points; If the target path point exists, then the target path point is removed from the initial path points to obtain the updated path point; The optimal driving path is obtained by optimizing the path based on the path distance between path points after the update, with the goal of minimizing the total path distance.

7. An IGV intelligent autonomous driving system based on multi-mode navigation, characterized in that, The method is applied to the IGV intelligent autonomous driving system based on multi-mode navigation as described in any one of claims 1 to 6; the IGV intelligent autonomous driving system based on multi-mode navigation includes: The data acquisition module is used to collect the first perception data, second perception data and third perception data of the intelligent guided vehicle based on the lidar sensor, vision sensor and inertial measurement sensor respectively; The data detection module is used to perform validity detection on the first sensing data, the second sensing data, and the third sensing data, and obtain the validity detection results of each sensing data. The driving route generation module is used to combine the valid perception data determined by the validity detection results of each perception data into the input information of the current navigation mode to generate candidate driving routes; The driving path optimization module is used to optimize the candidate driving path based on the local environmental obstacle map constructed from the first perception data to obtain the optimal driving path. The driving control module is used to control the intelligent guided vehicle to drive automatically based on the optimal driving path.

8. An electronic device, comprising: Memory, used to store computer software programs; A processor for reading and executing the computer software program, characterized in that, when the processor executes the computer software program, it implements the IGV intelligent autonomous driving method based on multi-mode navigation as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium, wherein a computer software program is stored therein, characterized in that, When the computer software program is executed by the processor, it implements the IGV intelligent autonomous driving method based on multi-mode navigation as described in any one of claims 1 to 6.

Images