Path planning method and system for intelligent inspection bionic robot

Abstract

The invention relates to the technical field of robots, and particularly discloses an intelligent inspection bionic robot path planning method and system. According to the method, the inspection path fork information and the destination position of the intelligent inspection biomimetic robot are firstly obtained, then the multiple fork track information is obtained according to the inspection path fork information, and then the dynamic obstacle information and the static obstacle information of the obstacles encountered by the intelligent inspection biomimetic robot are obtained; the second driving delay time is acquired according to the dynamic obstacle information, the third driving delay time is acquired according to the static obstacle information, and finally the optimal path is acquired according to the second driving delay time and the third driving delay time, so that the time cost of different path branches encountering various obstacles can be comprehensively evaluated. And the path with the minimum total delay time is selected as the optimal path, so that the intelligent inspection bionic robot can be ensured to complete the task in a shorter time or complete the plan in a time closer to the preset inspection time.

Description

Technical Field

[0001] The present invention relates to the field of robot technology, and in particular to a method and system for intelligent inspection bionic robot path planning. Background Art

[0002] With the continuous advancement of science and technology, intelligent robots have been widely used in industrial inspection, logistics and transportation, security monitoring and other fields. Especially in complex or dangerous environments, such as oil and gas pipelines, power equipment, mines and other fields, inspection robots need to be able to complete inspection tasks quickly and accurately.

[0003] The existing inspection robot path planning method is mainly based on a pre-set path for inspection, but the pre-set path will change due to changes in the environment, which also requires waiting for the path to be set again. This method causes inspection delays for the inspection robot, which in turn causes the inspection robot to be unable to patrol according to the preset time. Therefore, an intelligent inspection bionic robot path planning method is needed to solve the above problem. Summary of the invention

[0004] The purpose of the present invention is to provide a path planning method and system for an intelligent inspection bionic robot to solve the technical problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions: A method for intelligent inspection bionic robot path planning, comprising: Obtaining inspection path information of the intelligent inspection bionic robot, wherein the inspection path information includes inspection path fork information and destination location; Acquire multiple path fork node positions according to the inspection path fork information, and acquire multiple path fork trajectory information from the multiple path fork node positions to the destination position; Obtaining obstacle feature information of obstacles encountered by the intelligent inspection bionic robot, wherein the obstacle feature information includes dynamic obstacle information and static obstacle information; Acquire the moving trajectory information of the dynamic obstacle according to the dynamic obstacle information, and traverse the coincidence points of the moving trajectory information and multiple path fork node positions to obtain multiple first coincidence path fork node positions; Acquire a plurality of corresponding first travel delay times according to the plurality of first coincident path fork node positions, and filter the plurality of first travel delay times according to the minimum time to obtain a second travel delay time; Acquire a fixed position of a static obstacle according to the static obstacle information, and traverse the coincidence points of the fixed position and a plurality of path fork node positions to obtain a second coincidence path fork node position; Acquire a corresponding third travel delay time according to the fork node position of the second coincident path; A preferred path is acquired according to the second travel delay time and the third travel delay time, and path planning is performed for the intelligent inspection bionic robot according to the preferred path.

[0006] Preferably, the step of acquiring the positions of multiple path fork nodes according to the inspection path fork information comprises: Obtaining the main inspection path information of the intelligent inspection bionic robot according to the inspection path information, and obtaining corresponding path continuous points according to the inspection main path information; Acquire multiple branch path node information according to the inspection path fork information, and acquire corresponding multiple branch path terminal positions according to the multiple branch path node information; Acquire multiple branch path end point positions merged into multiple point positions of the path continuous points; The corresponding path fork starting points are obtained according to the multiple merging point positions, and the multiple path fork starting points are used as multiple path fork node positions.

[0007] Preferably, the step of traversing the coincidence points of the movement trajectory information and the positions of multiple path fork nodes to obtain multiple first coincidence path fork node positions comprises: Acquire the moving speed and moving direction of the dynamic obstacle according to the moving trajectory information, and acquire the moving trajectory curve of the dynamic obstacle within a preset time according to the moving speed and moving direction; Mapping the movement trajectory curve and the positions of multiple path fork nodes into a preset grid map to obtain multiple movement trajectory coordinate points and multiple path fork node coordinate points; Calculate multiple Euclidean distances based on each of the moving trajectory coordinate points and multiple path fork node coordinate points; Determine in sequence whether a plurality of the Euclidean distances are less than a preset threshold; If the Euclidean distance is less than a preset threshold, the moving trajectory coordinate points corresponding to the Euclidean distance that meets the condition are used as a plurality of first coincident path fork node positions that coincide with the path fork node coordinate points.

[0008] Preferably, the step of acquiring corresponding multiple first travel delay times according to multiple first coincident path fork node positions comprises: Acquire path capacity according to the position of the fork node of the first coincident path, wherein the path capacity represents the maximum passing volume; Acquire multiple collision volumes at the locations of the fork nodes of the first coincident path based on radar detection; Acquire path flow according to the plurality of collision volumes and preset elapsed time; The path flow density is calculated according to the path capacity and the path flow, wherein the calculation formula is: ; Among them, Q(M) represents the path flow density, L(j) represents the path flow, and L(l) represents the path capacity; Acquire a first travel delay time according to the path flow density; Repeat the steps of obtaining the path capacity according to the first overlapping path fork node position to obtaining the first travel delay time according to the path flow density to traverse all the first overlapping path fork node positions to obtain multiple first travel delay times.

[0009] Preferably, the step of acquiring the preferred route according to the second travel delay time and the third travel delay time comprises: Acquire a corresponding first preferred fork path according to the second travel delay time; Acquire a corresponding second preferred fork path according to the third travel delay time; Obtain the preset inspection path of the intelligent inspection bionic machine; Mapping the first preferred fork path, the second preferred fork path and the preset inspection path into a two-dimensional coordinate system for splicing to obtain a spliced ​​path; The splicing path is taken as the preferred path.

[0010] Preferably, the step of performing path planning for the intelligent inspection bionic robot according to the preferred path comprises: Acquire a preferred path driving trajectory according to the preferred path; Acquire multiple turning angles according to the path driving trajectory; Determining whether the plurality of turning angles are less than a preset angle; If the turning angle is smaller than the preset angle, the turning angle smaller than the preset angle is smoothly fitted to obtain a first fitting point and a second fitting point; Inserting a smooth node between the first fitting point and the second fitting point to generate a smooth turning angle, and optimizing the turning angle of the selected path driving trajectory according to the smooth turning angle to obtain a smooth preferred path; The path planning of the intelligent inspection bionic robot is carried out according to the smooth optimal path.

[0011] The present invention also provides an intelligent inspection bionic robot path planning system, comprising: A first acquisition module is used to acquire the inspection path information of the intelligent inspection bionic robot, wherein the inspection path information includes the inspection path fork information and the destination location; A second acquisition module is used to acquire multiple path fork node positions according to the inspection path fork information, and acquire multiple fork trajectory information from the multiple path fork node positions to the destination position; A third acquisition module is used to obtain obstacle feature information of obstacles encountered by the intelligent inspection bionic robot, wherein the obstacle feature information includes dynamic obstacle information and static obstacle information; A fourth acquisition module is used to acquire the moving trajectory information of the dynamic obstacle according to the dynamic obstacle information, and traverse the coincidence points of the moving trajectory information and a plurality of path fork node positions to obtain a plurality of first coincidence path fork node positions; A fifth acquisition module, configured to acquire a plurality of corresponding first travel delay times according to a plurality of first coincident path fork node positions, and filter the plurality of first travel delay times according to a minimum time to obtain a second travel delay time; A sixth acquisition module, configured to acquire a fixed position of a static obstacle according to the static obstacle information, and traverse the coincidence points of the fixed position and a plurality of path fork node positions to obtain a second coincidence path fork node position; A seventh acquisition module, configured to acquire a corresponding third travel delay time according to the fork node position of the second coincident path; An eighth acquisition module is used to acquire a preferred path according to the second driving delay time and the third driving delay time, and to perform path planning for the intelligent inspection bionic robot according to the preferred path.

[0012] Preferably, the second acquisition module includes: A first replacement unit is used to obtain the main inspection path information of the intelligent inspection bionic robot according to the inspection path information, and obtain the corresponding path continuous points according to the inspection main path information; A second replacement unit is used to obtain a plurality of branch path node information according to the inspection path fork information, and obtain a plurality of corresponding branch path terminal positions according to the plurality of branch path node information; A third replacement unit is used to obtain a plurality of merging point positions where the end points of the branch paths are merged into the continuous points of the path; The fourth replacement unit is used to obtain corresponding path fork starting points according to the multiple merging point positions, and use the multiple path fork starting points as multiple path fork node positions.

[0013] The present application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0014] The present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the above method are implemented.

[0015] The beneficial effects of the present application are as follows: the present application first obtains the inspection path information of the intelligent inspection bionic robot, wherein the inspection path information includes the inspection path fork information and the destination position, then obtains multiple fork trajectory information according to the inspection path fork information, and then obtains the obstacle feature information of the obstacle encountered by the intelligent inspection bionic robot, wherein the obstacle feature information includes dynamic obstacle information and static obstacle information, obtains the second travel delay time according to the dynamic obstacle information, and obtains the third travel delay time according to the static obstacle information, finally, obtains the preferred path according to the second travel delay time and the third travel delay time, and plans the path of the intelligent inspection bionic robot according to the preferred path, so that by combining the second travel delay time caused by dynamic obstacles and the third travel delay time caused by static obstacles, it is possible to comprehensively evaluate the time cost of different path branches when encountering various obstacles, and select the path with the smallest total delay time as the preferred path, so that the robot can minimize the time loss caused by obstacles during the inspection process, and there is no need to wait for the path to be set, thereby ensuring that the intelligent inspection bionic robot completes the task in a shorter time or completes the plan closer to the preset inspection time. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The figure is a schematic diagram of a method flow of an embodiment of the present application.

[0017] Figure 2 A schematic diagram of the system structure of an embodiment of the present application.

[0018] Figure 3 A schematic diagram of the internal structure of a computer device according to an embodiment of the present application.

[0019] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0020] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0021] like Figure 1-3 As shown, the present application provides a path planning method for an intelligent inspection bionic robot, comprising: S1. Obtaining inspection path information of the intelligent inspection bionic robot, wherein the inspection path information includes inspection path fork information and destination location; S2. Acquire multiple path fork node positions according to the inspection path fork information, and acquire multiple path fork trajectory information from the multiple path fork node positions to the destination position; S3, obtaining obstacle feature information of obstacles encountered by the intelligent inspection bionic robot, wherein the obstacle feature information includes dynamic obstacle information and static obstacle information; S4. Acquire the moving trajectory information of the dynamic obstacle according to the dynamic obstacle information, and traverse the coincidence points of the moving trajectory information and multiple path fork node positions to obtain multiple first coincidence path fork node positions; S5, acquiring corresponding multiple first travel delay times according to multiple first coincident path fork node positions, and filtering the multiple first travel delay times according to the minimum time to obtain a second travel delay time; S6. Obtain a fixed position of the static obstacle according to the static obstacle information, and traverse the coincidence points of the fixed position and a plurality of path fork node positions to obtain a second coincidence path fork node position; S7, acquiring a corresponding third travel delay time according to the fork node position of the second overlapping path; S8. Obtain a preferred path according to the second travel delay time and the third travel delay time, and perform path planning for the intelligent inspection bionic robot according to the preferred path.

[0022] As described in the above steps S1-S8, the existing inspection robot path planning method is mainly based on a pre-set path for inspection, but the pre-set path will change due to changes in the environment, which also leads to the need to wait for the path to be set again. This method causes inspection delays for the inspection robot, which in turn causes the inspection robot to be unable to patrol according to the preset time. Therefore, the present invention first obtains the inspection path information of the intelligent inspection bionic robot, wherein the inspection path information includes the inspection path fork information and the destination location, so that the inspection path information can provide a decision-making basis for the subsequent inspection robot path planning, and then obtains multiple path fork node positions according to the inspection path fork information, and obtains multiple path fork node positions The information of multiple forks in the trajectory to the destination location, where the fork trajectory information is an alternative path on the pre-set path, and its role is to avoid obstacles. This refines the intermediate links of path planning and converts the abstract inspection path information into specific and operational path fork nodes and trajectory information, which enables subsequent calculations and decisions to be based on more accurate path data, improves the accuracy and effectiveness of path planning, and also provides basic selection conditions for the screening of paths after encountering obstacles. Then, the obstacle feature information of the intelligent inspection bionic robot is obtained, where the obstacle feature information includes dynamic obstacle information and static obstacle information. Such real-time acquisition of obstacle information is the key to the robot's safe and efficient inspection in complex environments. The acquisition of dynamic obstacle information and static obstacle information enables the robot to fully understand the obstacles in the environment, providing a basis for subsequent obstacle avoidance and delay reduction. For example, knowing that there are dynamic pedestrians or vehicles ahead, as well as fixed obstacles such as pillars or walls, the robot can plan a detour route in advance, and also provide a basis for path optimization. Since the movement of dynamic obstacles increases the complexity of path planning, it is necessary to predict its impact on the robot path. Therefore, the movement trajectory information of the dynamic obstacle is obtained according to the dynamic obstacle information, and the movement trajectory information and multiple path fork node positions are traversed at the coincidence point to obtain multiple first coincident path fork node positions. In this way, by analyzing the relationship between the movement trajectory of the dynamic obstacle and the path fork node position, the fork nodes that may coincide with the dynamic obstacle trajectory are determined. These nodes are potential candidate path nodes, and early identification helps to calculate the delay time and select the optimal path in the subsequent process. For example, if the trajectory of the dynamic obstacle coincides with the position of the fork node D, the robot needs to consider the possible delay caused by passing through point D when planning the path. At the same time, through this coincidence point traversal method, the key nodes affected by the dynamic obstacle can be accurately found, providing targeted data for the subsequent delay time calculation and path optimization.If this step is not performed, the robot may blindly choose a path when facing a dynamic obstacle, resulting in a greater delay. Then, the corresponding multiple first travel delay times are obtained according to the multiple first coincident path fork node positions, and the multiple first travel delay times are screened according to the minimum time to obtain the second travel delay time. The travel delay time of each fork node that may be affected by a dynamic obstacle is calculated, reflecting the time loss that the robot may encounter at these nodes. Screening out the minimum delay time helps to determine the relatively optimal path branch to reduce the impact of dynamic obstacles on inspection efficiency. For example, after calculation, the first travel delay time of fork node E is 5 seconds, and that of fork node F is 3 seconds. Then, selecting node F may shorten the overall inspection time. At the same time, multiple first travel delay times can be used to quantify the index of path optimization, and can provide data support for the robot to inspect on time. Then, the fixed position of the static obstacle is obtained according to the static obstacle information, and the fixed position and multiple path fork node positions are traversed to obtain the second coincident path fork node position. Among them, the processing method of static obstacle information is similar to that of dynamic obstacles. The fork nodes that coincide with the static obstacle positions are determined. These nodes will also affect the robot's travel and may require detours or adjustments to the path, providing a basis for the subsequent calculation of the delay time caused by static obstacles. At the same time, by traversing the coincidence points to find the relevant fork nodes, the impact of static obstacles on the path can be more targeted. Provide complete data support for comprehensively considering the impact of dynamic and static obstacles and selecting the final preferred path. If the static obstacle is ignored, the robot may directly collide with the obstacle and fail to complete the task. Then, the corresponding third travel delay time is obtained according to the position of the fork node of the second overlapping path. In this way, the impact of the static obstacle on the robot needs to be specifically quantified. The calculation of the third travel delay time can accurately reflect its impact on the path efficiency. Only by obtaining the respective and combined delay times of the dynamic and static obstacles can the optimal path be scientifically selected in the subsequent steps to ensure that the robot can complete the inspection task within the preset time during the inspection process. Finally, the preferred path is obtained according to the second travel delay time and the third travel delay time, and the intelligent inspection bionic robot is path planned according to the preferred path. In this way, by combining the second travel delay time caused by the dynamic obstacle and the third travel delay time caused by the static obstacle, the time cost of different path branches when encountering various obstacles can be comprehensively evaluated, and the path with the smallest total delay time is selected as the preferred path, which can minimize the time loss caused by obstacles during the inspection process. At the same time, there is no need to wait for the setting path, thereby ensuring that the intelligent inspection bionic robot completes the task in a shorter time or completes the plan closer to the preset inspection time.

[0023] In one embodiment, the step S2 of acquiring the positions of multiple path fork nodes according to the inspection path fork information includes: S201, obtaining the main inspection path information of the intelligent inspection bionic robot according to the inspection path information, and obtaining corresponding path continuous points according to the inspection main path information; S202, obtaining multiple branch path node information according to the inspection path fork information, and obtaining multiple corresponding branch path terminal positions according to the multiple branch path node information; S203, acquiring multiple branch path end point positions merged into multiple path continuous point positions; S204: Obtain corresponding path fork starting points according to the plurality of merging point positions, and use the plurality of path fork starting points as a plurality of path fork node positions.

[0024] As described in the above steps S201-S204, the present invention first obtains the main inspection path information of the intelligent inspection bionic robot based on the inspection path information, and obtains the corresponding path continuous points based on the inspection main path information. Obtaining the inspection main path information in this way helps to determine the main travel direction and route framework of the robot in the entire inspection area, which is the basis for building a complete path planning. In a complex inspection environment, first determining the main path is the key to effective path planning. The main path information provides a basic direction guide for the entire path planning, so that the robot can have a main travel direction as a reference when faced with many possible path choices, avoiding confusion and disorder in path planning. At the same time, it is not enough to only know the general direction of the main path. The path continuous points can more accurately portray the actual form of the main path. In the subsequent operations such as calculating the path length and judging the positional relationship with other nodes or obstacles, accurate path description is indispensable, which can improve the accuracy and reliability of path planning. In actual inspection scenarios, the path is often not a single straight line, but has multiple branches. In order to enable the robot to make a reasonable path selection at the fork, multiple branch path node information is obtained according to the inspection path fork information, wherein the branch path node information is the secondary bifurcation path of the inspection path fork information. Only two bifurcation situations are listed in the present invention, but the number of times the path can be bifurcated does not affect the path selection, because the present invention only selects the point position connected to the continuous point position of the path, and other bifurcated paths that are not connected are invalid paths, and the corresponding multiple branch path terminal positions are obtained according to the multiple branch path node information, and obtaining the branch path node information can clarify the possible branch positions on the main path. These nodes are the key decision points that the robot needs to make path selection during the inspection process. At the same time, understanding the position of the branch path node helps the robot prepare in advance to deal with different path selection situations, and obtaining the branch path terminal position provides the robot with a potential target range for each branch path. This allows the robot to have a preliminary understanding of the destination and possible areas of each branch path when considering path selection, so as to better evaluate the impact of different branch paths on the completion of inspection tasks. In order to achieve efficient path planning, the main path and branch paths need to be considered as a whole. By incorporating the merging point positions, a unified model containing all key path nodes can be constructed. Therefore, it is necessary to obtain multiple branch path end positions that are merged into the multiple merging point positions of the path continuous points. In this way, the branch path end positions are merged into the path continuous points, which can organically combine the information of the main path and the branch path to form a complete path network representation.In this way, the robot can perform path planning and decision-making in a unified path system, and can improve the integrity and coherence of path planning. Finally, the corresponding path fork starting point is obtained according to the multiple point positions, wherein the path fork starting point is reversely deduced through the point position, and its purpose is to eliminate invalid forks, and the multiple path fork starting points are used as multiple path fork node positions, and the path fork starting point is the position where the robot actually faces path selection in actual inspection. It is determined as the path fork node position, so that the robot can clearly know at which specific points decisions need to be made. These node positions become the core reference points for subsequent calculation of path trajectory, evaluation of obstacle impact, and selection of the optimal path.

[0025] In one embodiment, the step S4 of traversing the coincidence points of the movement trajectory information and the plurality of path fork node positions to obtain a plurality of first coincidence path fork node positions includes: S401, obtaining a moving speed and a moving direction of a dynamic obstacle according to the moving trajectory information, and obtaining a moving trajectory curve of the dynamic obstacle within a preset time according to the moving speed and the moving direction; S402, mapping the movement trajectory curve and the positions of multiple path fork nodes to a preset grid map to obtain multiple movement trajectory coordinate points and multiple path fork node coordinate points; S403, calculating a plurality of Euclidean distances according to each of the moving trajectory coordinate points and a plurality of the path fork node coordinate points; S404, determining in sequence whether the plurality of Euclidean distances are less than a preset threshold; If the Euclidean distance is less than a preset threshold, the moving trajectory coordinate points corresponding to the Euclidean distance that meets the condition are used as a plurality of first coincident path fork node positions that coincide with the path fork node coordinate points.

[0026] As described in the above steps S401-S404, the present invention first obtains the moving speed and moving direction of the dynamic obstacle according to the moving trajectory information, and obtains the moving trajectory curve of the dynamic obstacle within a preset time according to the moving speed and moving direction. The moving trajectory curve within the preset time is calculated by the moving speed and direction, and the robot can know in advance the approximate position range of the dynamic obstacle in the future. This helps the robot to consider the impact of dynamic obstacles in advance during path planning, avoid collisions with them, and improve the safety and fluency of the inspection process. Then, the moving trajectory curve and the positions of multiple path forks are mapped to the preset grid map to obtain multiple moving trajectory coordinate points and multiple path fork node coordinate points, and the moving trajectory curve and the path fork node position are mapped to the preset grid map, providing a unified spatial reference system for the two. In this unified coordinate system, the robot can easily compare and analyze the positions of dynamic obstacles and path nodes, which is convenient for subsequent calculation of the relationship between them, and at the same time, the moving trajectory coordinate points and the path fork node coordinate points are obtained, so that the positions of dynamic obstacles and path node positions can be represented on the map in the form of precise coordinates. This helps the robot to more accurately determine the relative positional relationship between the dynamic obstacle and the path fork node, and provides an accurate data basis for operations such as calculating the Euclidean distance. Then, multiple Euclidean distances are calculated based on each of the moving trajectory coordinate points and multiple path fork node coordinate points. Calculating the Euclidean distance can accurately quantify the proximity between each moving trajectory coordinate point and the path fork node coordinate point. Through the distance value, the robot can intuitively understand the distance relationship between the dynamic obstacle and each path fork node during the movement process, and then judge in turn whether the multiple Euclidean distances are less than the preset threshold. If the Euclidean distance is less than the preset threshold, the moving trajectory coordinate point corresponding to the Euclidean distance that meets the condition is used as the multiple first coincident path fork node positions that coincide with the path fork node coordinate point. By comparing with the preset threshold, it can accurately determine whether the moving trajectory coordinate point is close enough to the path fork node coordinate point, thereby determining whether the two coincide. Finding the first coincident path fork node position means that these nodes are key positions where dynamic obstacles may affect the robot's path selection, and provide clear target nodes for subsequent operations such as calculating delay time.

[0027] In one embodiment, the step S5 of acquiring corresponding multiple first travel delay times according to multiple first coincident path fork node positions includes: S501, obtaining a path capacity according to the position of the fork node of the first coincident path, wherein the path capacity represents a maximum passing volume; S502, acquiring multiple collision volumes at the locations of the fork nodes of the first overlapping paths based on radar detection; S504, obtaining a path flow rate according to the plurality of collision volumes and a preset elapsed time; S505: Calculate the path flow density according to the path capacity and the path flow, wherein the calculation formula is: ; Among them, Q(M) represents the path flow density, L(j) represents the path flow, and L(l) represents the path capacity; S506. Obtain a first travel delay time according to the path flow density, and repeat the steps of obtaining the path capacity according to the first overlapping path fork node position to obtaining the first travel delay time according to the path flow density, so as to traverse all first overlapping path fork node positions to obtain multiple first travel delay times.

[0028] As described in the above steps S501-S506, the present invention first obtains the path capacity according to the position of the fork node of the first overlapping path, wherein the path capacity represents the maximum passing volume, and the path capacity is an indicator of the ability of the path to accommodate objects in the physical space. Obtaining the path capacity provides basic data for the subsequent analysis of the passage of the path under the influence of dynamic obstacles, which helps to determine whether the path will become congested or impassable due to the existence of dynamic obstacles. Then, based on radar detection, multiple collision volumes at the position of the fork node of the first overlapping path are obtained. In this way, the collision volume is obtained by radar detection, which can accurately represent the size of the space occupied by the dynamic obstacle at the position of the fork node of the first overlapping path. This helps the robot to more intuitively understand the occupation of the path space by the dynamic obstacle, and provides key data for evaluating the degree of obstruction to the passage of the path. Then, the path flow is obtained according to the multiple collision volumes and the preset elapsed time, and then the path flow density is calculated according to the path capacity and the path flow, and the path flow density comprehensively calculates the path capacity and the path flow to obtain a standardized index for evaluating the passage efficiency of the path under the influence of dynamic obstacles. It can intuitively reflect the degree of congestion of the path. The larger the value, the more congested the path is, and the greater the delay that the robot may encounter when passing through. Finally, the first travel delay time is obtained according to the path flow density, and the steps of obtaining the path capacity according to the first coincident path fork node position to obtaining the first travel delay time according to the path flow density are repeated to traverse all the first coincident path fork node positions to obtain multiple first travel delay times. In this way, the first travel delay time is calculated according to the path flow density, which directly reflects the time delay caused by the dynamic obstacle at the first coincident path fork node position to the robot's travel. By obtaining the travel delay time of multiple first coincident path fork nodes, the robot can fully understand the delay at different nodes that may be affected by dynamic obstacles, and provide accurate data support for selecting the optimal path. At the same time, the calculation steps are repeated for multiple first coincident path fork node positions, which can cover the dynamic obstacle influence that the robot may encounter under different path branch selections. This enables the robot to comprehensively consider various possible scenarios when planning the path, select the path with the smallest overall delay, and improve the comprehensiveness and adaptability of path planning.

[0029] In one embodiment, the step S8 of acquiring the preferred path according to the second travel delay time and the third travel delay time includes: S801, acquiring a corresponding first preferred fork path according to the second travel delay time; S802, acquiring a corresponding second preferred fork path according to the third travel delay time; S803, obtaining a preset inspection path of the intelligent inspection bionic machine; S804, mapping the first preferred fork path, the second preferred fork path and the preset inspection path into a two-dimensional coordinate system for splicing to obtain a spliced ​​path; The splicing path is taken as the preferred path.

[0030] As described in the above steps S801-S804, since the movement of dynamic obstacles is uncertain, it has a great impact on the path planning of the robot. Determining the preferred fork path based solely on the delay time caused by dynamic obstacles can more focusedly deal with the challenges brought by dynamic obstacles, and provide a path basis optimized based on dynamic conditions for the subsequent comprehensive consideration of the impact of static obstacles, so that the path planning is more hierarchical and targeted. On this basis, the present invention first obtains the corresponding first preferred fork path according to the second travel delay time, and the second travel delay time reflects the degree of influence of dynamic obstacles on each path fork node. By selecting the first preferred fork path based on this time, the robot can give priority to the path branch with the smallest delay under the influence of dynamic obstacles, thereby avoiding the greater impact of dynamic obstacles on the inspection efficiency to a certain extent, ensuring that it can move to the destination more quickly in a dynamic environment, and then obtain the corresponding second preferred fork path according to the third travel delay time. Different from the first preferred fork path based on dynamic obstacles, the second preferred fork path focuses on static obstacle factors. The combination of the two can provide a more comprehensive basis for path selection for the robot, so that the robot can cope with dynamic changes and fully consider static environmental factors when planning the path, so as to develop a more reasonable and efficient inspection path, and then obtain the preset inspection path of the intelligent inspection bionic machine. The preset inspection path is a basic inspection route framework set based on prior knowledge or conventional task requirements. It provides an initial reference direction for the robot's path planning, so that the robot can make a comprehensive decision in combination with the preset path while considering the impact of dynamic and static obstacles, so as to avoid the path planning from completely deviating from the expected goal and ensure that the inspection task can cover the necessary area. Then, the first preferred fork path, the second preferred fork path and the preset inspection path are mapped to the two-dimensional coordinate system for splicing to obtain the spliced ​​path. By splicing the first preferred fork path based on dynamic obstacles, the second preferred fork path based on static obstacles and the preset inspection path in the two-dimensional coordinate system, a comprehensive and integrated consideration of the impact of dynamic and static obstacles and the preset task requirements is achieved. The obtained splicing path can avoid obstacles and reduce delay time to the greatest extent, while following the preset inspection logic, thereby determining an overall optimal inspection path and improving the robot's inspection efficiency and task completion quality.

[0031] In one embodiment, the step S8 of performing path planning for the intelligent inspection bionic robot according to the preferred path includes: S805, obtaining a preferred path driving trajectory according to the preferred path; S806, obtaining a plurality of turning angles according to the path driving trajectory; S807, determining whether the plurality of turning angles are less than a preset angle; If the turning angle is smaller than the preset angle, the turning angle smaller than the preset angle is smoothly fitted to obtain a first fitting point and a second fitting point; S808, inserting a smooth node between the first fitting point and the second fitting point to generate a smooth turning angle, and optimizing the turning angle of the selected path driving trajectory according to the smooth turning angle to obtain a smooth preferred path; S809: Perform path planning for the intelligent inspection bionic robot according to the smooth optimal path.

[0032] As described in the above steps S805-S809, since the ultimate goal of path planning is to enable the robot to actually move along the planned path. Acquiring the preferred path driving trajectory establishes a connection between path planning and robot motion execution, converting the abstract path selection into a specific motion instruction sequence that the robot can understand and execute, which is a key transition link from planning to execution in the entire inspection process. Therefore, the present invention first obtains the preferred path driving trajectory according to the preferred path, and obtaining the preferred path driving trajectory can express the previously determined preferred path in the form of specific geometric lines, including the starting point, end point, turning point of the path, and the connection method of each line segment. This allows the robot to have a more intuitive and accurate understanding of the actual route to be traveled, and provides basic data for the subsequent analysis of the geometric characteristics of the path (such as the turning angle), and then obtains multiple turning angles according to the path driving trajectory, and the turning angle is an important indicator for measuring the complexity of the path. By obtaining multiple turning angles, the robot can understand how many turning points there are in the preferred path and the degree of curvature of each turning point. A larger turning angle or more turning points may mean a more complicated path. The robot needs more adjustments and controls during driving, which places higher requirements on the robot's maneuverability and stability. At the same time, it reduces the energy consumption, time delay or even collision risk caused by sharp turns, improves inspection efficiency and safety, and determines whether the multiple turning angles are less than the preset angle. If the turning angle is less than the preset angle, the turning angle less than the preset angle is smoothly fitted to obtain a first fitting point and a second fitting point. Smooth fitting is a mathematical method used to find a smooth curve or function to approximate these data points given a set of data points (points on the path in the present invention), so that the curve approximates these data points in the data. The transition between the points is natural, without sharp changes or mutations, thereby reducing the tortuosity of the path, allowing the robot to turn more smoothly during driving, reducing energy consumption and steering difficulty, and improving driving speed and stability. In the smooth fitting process, the fitting point refers to the data point used to construct a smooth curve or function. In the present invention, when it is judged that the turning angle is less than the preset angle, the path points near these turning points will be processed, and the points obtained for determining the shape of the smooth curve are the fitting points. The first fitting point and the second fitting point are the points on both sides of the turning angle. They are the key positions where the smooth curve passes, and are used to determine how to insert smooth nodes at the turning points of the original path to generate a smooth turning angle, thereby optimizing the geometric shape of the entire path. The determination and processing of the fitting points are important steps to achieve path smoothing optimization, which directly affects the quality of the final smooth preferred path and the driving performance of the robot. The setting of the preset angle provides a standard for determining whether the turning angle needs to be optimized.By comparing with the preset angle, the robot can quickly identify which turning angles are large and may need to be optimized, and which turning angles are within an acceptable range and do not require additional adjustments. This helps to focus on optimizing key turning points and improve the efficiency of path optimization. At the same time, by judging whether the turning angle is less than the preset angle, the robot can avoid encountering turns that are too sharp or difficult to handle during driving, reduce the risks of collision, instability, etc. caused by improper turning, and ensure the smooth progress of the inspection task. The turning angles that are less than the preset angle are smoothed and smooth nodes are inserted, which can make the path at the turning point smoother and reduce the tortuosity of the path. In this way, the steering difficulty and energy consumption of the robot when turning can be reduced, and the speed and stability of the robot passing the turning point can be improved, thereby optimizing the geometry of the entire path and making the robot travel more smoothly. Then, a smooth node is inserted between the first fitting point and the second fitting point to generate a smooth turning angle, and the selected path driving trajectory is optimized according to the smooth turning angle to obtain a smooth preferred path, and the smooth preferred path can make the path at the turning point smoother and reduce the tortuosity of the path. This can reduce the difficulty and energy consumption of the robot when turning, improve the speed and stability of the robot passing the turning point, thereby optimizing the geometry of the entire path and making the robot travel more smoothly. Finally, the intelligent inspection bionic robot is planned according to the smooth optimal path. After a series of calculations and optimizations, the smooth optimal path becomes the final path for the robot to actually perform the inspection task. It comprehensively considers factors such as the impact of obstacles and the optimization of the path geometry, and provides the robot with a driving route that best suits the current environment and task requirements, ensuring that the robot can complete the inspection task efficiently, safely and accurately.

[0033] The present application also provides an intelligent inspection bionic robot path planning system, comprising: The first acquisition module 1 is used to obtain the inspection path information of the intelligent inspection bionic robot, wherein the inspection path information includes the inspection path fork information and the destination location; A second acquisition module 2 is used to acquire multiple path fork node positions according to the inspection path fork information, and acquire multiple path fork trajectory information from the multiple path fork node positions to the destination position; The third acquisition module 3 is used to obtain obstacle feature information of obstacles encountered by the intelligent inspection bionic robot, wherein the obstacle feature information includes dynamic obstacle information and static obstacle information; A fourth acquisition module 4 is used to acquire the moving trajectory information of the dynamic obstacle according to the dynamic obstacle information, and traverse the coincidence points of the moving trajectory information and a plurality of path fork node positions to obtain a plurality of first coincidence path fork node positions; A fifth acquisition module 5, configured to acquire a plurality of corresponding first travel delay times according to a plurality of first coincident path fork node positions, and filter the plurality of first travel delay times according to a minimum time to obtain a second travel delay time; A sixth acquisition module 6 is used to acquire a fixed position of a static obstacle according to the static obstacle information, and perform coincidence traversal on the fixed position and a plurality of path fork node positions to obtain a second coincidence path fork node position; A seventh acquisition module 7, configured to acquire a corresponding third travel delay time according to the fork node position of the second coincident path; The eighth acquisition module 8 is used to acquire a preferred path according to the second travel delay time and the third travel delay time, and perform path planning for the intelligent inspection bionic robot according to the preferred path.

[0034] In one embodiment, the second acquisition module 2 includes: A first replacement unit is used to obtain the main inspection path information of the intelligent inspection bionic robot according to the inspection path information, and obtain the corresponding path continuous points according to the inspection main path information; A second replacement unit is used to obtain a plurality of branch path node information according to the inspection path fork information, and obtain a plurality of corresponding branch path terminal positions according to the plurality of branch path node information; A third replacement unit is used to obtain a plurality of merging point positions where the end points of the branch paths are merged into the continuous points of the path; The fourth replacement unit is used to obtain corresponding path fork starting points according to the multiple merging point positions, and use the multiple path fork starting points as multiple path fork node positions.

[0035] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned deep learning-based PCB component detection method when executing the computer program.

[0036] The present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the above-mentioned method for detecting PCB components based on deep learning are implemented.

[0037] Those of ordinary skill in the art can understand that the implementation of all or part of the process in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the process of the embodiment of each method as described above. Among them, any reference to memory, storage, database or other media provided in this application and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0038] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also other elements not explicitly listed, or includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0039] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for intelligent inspection bionic robot path planning, characterized in that: include: Obtaining inspection path information of the intelligent inspection bionic robot, wherein the inspection path information includes inspection path fork information and destination location; Acquire multiple path fork node positions according to the inspection path fork information, and acquire multiple path fork trajectory information from the multiple path fork node positions to the destination position; Obtaining obstacle feature information of obstacles encountered by the intelligent inspection bionic robot, wherein the obstacle feature information includes dynamic obstacle information and static obstacle information; Acquire the moving trajectory information of the dynamic obstacle according to the dynamic obstacle information, and traverse the coincidence points of the moving trajectory information and multiple path fork node positions to obtain multiple first coincidence path fork node positions; Acquire a plurality of corresponding first travel delay times according to the plurality of first coincident path fork node positions, and filter the plurality of first travel delay times according to the minimum time to obtain a second travel delay time; Acquire a fixed position of a static obstacle according to the static obstacle information, and traverse the coincidence points of the fixed position and a plurality of path fork node positions to obtain a second coincidence path fork node position; Acquire a corresponding third travel delay time according to the fork node position of the second coincident path; A preferred path is acquired according to the second travel delay time and the third travel delay time, and path planning is performed for the intelligent inspection bionic robot according to the preferred path.

2. The intelligent inspection bionic robot path planning method according to claim 1 is characterized in that: The step of acquiring the positions of multiple path fork nodes according to the inspection path fork information comprises: Obtaining the main inspection path information of the intelligent inspection bionic robot according to the inspection path information, and obtaining corresponding path continuous points according to the inspection main path information; Acquire multiple branch path node information according to the inspection path fork information, and acquire corresponding multiple branch path terminal positions according to the multiple branch path node information; Acquire multiple branch path end point positions merged into multiple path continuous point positions; The corresponding path fork starting points are obtained according to the multiple merging point positions, and the multiple path fork starting points are used as multiple path fork node positions.

3. The intelligent inspection bionic robot path planning method according to claim 1, characterized in that: The step of traversing the coincidence points of the movement trajectory information and the positions of multiple path fork nodes to obtain multiple first coincidence path fork node positions includes: Acquire the moving speed and moving direction of the dynamic obstacle according to the moving trajectory information, and acquire the moving trajectory curve of the dynamic obstacle within a preset time according to the moving speed and moving direction; Mapping the movement trajectory curve and the positions of multiple path fork nodes into a preset grid map to obtain multiple movement trajectory coordinate points and multiple path fork node coordinate points; Calculate multiple Euclidean distances based on each of the moving trajectory coordinate points and multiple path fork node coordinate points; Determine in sequence whether a plurality of the Euclidean distances are less than a preset threshold; If the Euclidean distance is less than a preset threshold, the moving trajectory coordinate points corresponding to the Euclidean distance that meets the condition are used as a plurality of first coincident path fork node positions that coincide with the path fork node coordinate points.

4. The intelligent inspection bionic robot path planning method according to claim 1, characterized in that: The step of acquiring corresponding multiple first travel delay times according to multiple first coincident path fork node positions comprises: Acquire path capacity according to the position of the fork node of the first coincident path, wherein the path capacity represents the maximum passing volume; Acquire multiple collision volumes at the locations of the fork nodes of the first coincident path based on radar detection; Acquire path flow according to the plurality of collision volumes and preset elapsed time; The path flow density is calculated according to the path capacity and the path flow, wherein the calculation formula is: ； Among them, Q(M) represents the path flow density, L(j) represents the path flow, and L(l) represents the path capacity; Acquire a first travel delay time according to the path flow density; Repeat the steps of obtaining the path capacity according to the first overlapping path fork node position to obtaining the first travel delay time according to the path flow density to traverse all the first overlapping path fork node positions to obtain multiple first travel delay times.

5. The intelligent inspection bionic robot path planning method according to claim 1, characterized in that: The step of acquiring the preferred route according to the second travel delay time and the third travel delay time comprises: Acquire a corresponding first preferred fork path according to the second travel delay time; Acquire a corresponding second preferred fork path according to the third travel delay time; Obtain the preset inspection path of the intelligent inspection bionic machine; Mapping the first preferred fork path, the second preferred fork path and the preset inspection path into a two-dimensional coordinate system for splicing to obtain a spliced ​​path; The splicing path is taken as the preferred path.

6. The intelligent inspection bionic robot path planning method according to claim 1, characterized in that: The step of performing path planning for the intelligent inspection bionic robot according to the preferred path comprises: Acquire a preferred path driving trajectory according to the preferred path; Acquire multiple turning angles according to the path driving trajectory; Determining whether the plurality of turning angles are less than a preset angle; If the turning angle is smaller than the preset angle, the turning angle smaller than the preset angle is smoothly fitted to obtain a first fitting point and a second fitting point; Inserting a smooth node between the first fitting point and the second fitting point to generate a smooth turning angle, and optimizing the turning angle of the selected path driving trajectory according to the smooth turning angle to obtain a smooth preferred path; The path planning of the intelligent inspection bionic robot is carried out according to the smooth optimal path.

7. An intelligent inspection bionic robot path planning system, characterized in that: include: A first acquisition module is used to acquire the inspection path information of the intelligent inspection bionic robot, wherein the inspection path information includes the inspection path fork information and the destination location; A second acquisition module is used to acquire multiple path fork node positions according to the inspection path fork information, and acquire multiple fork trajectory information from the multiple path fork node positions to the destination position; A third acquisition module is used to obtain obstacle feature information of obstacles encountered by the intelligent inspection bionic robot, wherein the obstacle feature information includes dynamic obstacle information and static obstacle information; A fourth acquisition module is used to acquire the moving trajectory information of the dynamic obstacle according to the dynamic obstacle information, and traverse the coincidence points of the moving trajectory information and a plurality of path fork node positions to obtain a plurality of first coincidence path fork node positions; A fifth acquisition module, configured to acquire a plurality of corresponding first travel delay times according to a plurality of first coincident path fork node positions, and filter the plurality of first travel delay times according to a minimum time to obtain a second travel delay time; A sixth acquisition module, configured to acquire a fixed position of a static obstacle according to the static obstacle information, and traverse the coincidence points of the fixed position and a plurality of path fork node positions to obtain a second coincidence path fork node position; A seventh acquisition module, configured to acquire a corresponding third travel delay time according to the fork node position of the second coincident path; An eighth acquisition module is used to acquire a preferred path according to the second driving delay time and the third driving delay time, and to perform path planning for the intelligent inspection bionic robot according to the preferred path.

8. The intelligent inspection bionic robot path planning system according to claim 7, characterized in that: The second acquisition module includes: A first replacement unit is used to obtain the main inspection path information of the intelligent inspection bionic robot according to the inspection path information, and obtain the corresponding path continuous points according to the inspection main path information; A second replacement unit is used to obtain a plurality of branch path node information according to the inspection path fork information, and obtain a plurality of corresponding branch path terminal positions according to the plurality of branch path node information; A third replacement unit is used to obtain a plurality of merging point positions where the end points of the branch paths are merged into the continuous points of the path; The fourth replacement unit is used to obtain corresponding path fork starting points according to the multiple merging point positions, and use the multiple path fork starting points as multiple path fork node positions.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images