A path planning method suitable for rapid convergence of a mechanical arm in a complex environment

Abstract

This invention proposes a fast convergence path planning method suitable for robotic arms in complex environments. First, in collision detection, an obstacle expansion method is used to treat expanded obstacles as collision detection obstacles, preventing unnecessary potential collisions between the algorithm's execution unit and obstacles during physical experiments. Second, a better constraint sampling method is adopted to further narrow the sampling interval and sample within the restricted interval. Finally, a sampling point optimization strategy is introduced, designing a cost function and performing weight adaptation. The cost of multiple sampled candidate points is calculated, and the optimal point is selected as the final sampling point. This invention solves the problem of low path generation quality in complex environments for robotic arms.

Description

Technical Field

[0001] This invention relates to a path planning method for rapid convergence in complex environments for robotic arms. Background Technology

[0002] In key areas such as industrial automation, medical surgery, and warehousing and logistics, the operational accuracy and efficiency of robotic arms depend primarily on path planning technology. This technology not only ensures collision-free movement from the starting point to the target point but also considers multiple objectives, including shortest path, lowest energy consumption, and smooth motion. It is the core guarantee for the safe and efficient operation of robotic arms. With the deepening of Industry 4.0, the application environment of robotic arms is becoming increasingly complex. At the same time, robotic arms are developing towards multi-joint, high-degree-of-freedom models, with 6-axis and even 7-axis robotic arms becoming mainstream. This significantly increases the complexity of their kinematic models. Traditional path planning methods, due to their high computational cost and low efficiency, are unable to meet practical needs. Therefore, developing efficient planning technologies adapted to complex environments has become an industry necessity.

[0003] Over decades of evolution, robotic arm path planning has developed into a diverse algorithmic system, with each algorithm having its own strengths in terms of applicable scenarios and performance. In traditional algorithms, A... Graph search algorithms such as Dijkstra's algorithm find the optimal path by discretizing the workspace and relying on heuristic functions. They perform stably in simple structured environments (such as warehouse sorting with fixed shelves), but when faced with high-degree-of-freedom robotic arms, the number of discretized nodes increases exponentially, easily falling into the "curse of dimensionality". Geometric construction algorithms such as free space method and configuration space method can intuitively reflect the collision relationship between the robotic arm and the environment, but the difficulty of space construction in complex environments and the fact that the generated paths are mostly polylines that require additional smoothing processing increases the complexity of the planning process.

[0004] Intelligent optimization algorithms offer new solutions for complex scenarios. Genetic algorithms optimize paths by simulating the selection, crossover, and mutation processes of biological evolution, while particle swarm optimization (PSO) simulates bird flocks cooperating to update path solutions. Both do not require complex discretization, have strong global search capabilities, and are suitable for planning high-degree-of-freedom robotic arms. However, these algorithms also have significant drawbacks: genetic algorithms have slow convergence speeds and struggle to update paths in dynamic environments; PSO is sensitive to initial parameters, easily gets trapped in local optima, and is ill-suited for scenarios with frequently moving obstacles.

[0005] Current mainstream sampling algorithms strike a good balance between efficiency and applicability. RRT The algorithm is favored for its fast search speed, strong exploration ability, and probabilistic completeness, but it still has many unresolved problems, such as poor initial path quality, inefficient path generation, and failure to find feasible paths in complex environments. To address these issues... The problems with the algorithm have been the subject of in-depth research by many scholars. To address the issues in the original RRT... The RRT-connect algorithm addresses the problems of slow search speed and difficulty in connecting target regions in complex high-dimensional spaces. It introduces the idea of ​​bidirectional alternating tree expansion, significantly improving path generation success rate and path expansion speed. To solve the problem of poor initial path quality caused by the high randomness and non-optimal solutions of the original algorithm, RRT... The algorithm reduces randomness by deselecting parent nodes and reconnecting to parent nodes with lower costs, thus obtaining an asymptotically optimal path. (Informed-RRT) The algorithm reduces meaningless sampling by limiting the sampling interval, thereby accelerating convergence to the optimal path.

[0006] Although existing research has improved the effectiveness of RRT by addressing specific issues, and some studies have attempted to combine multiple methods to generate more comprehensive methods with significant results, they have not fully considered the applicability and effectiveness of the corresponding algorithms in complex 3D environments. Summary of the Invention

[0007] This invention proposes a path planning method that converges quickly in complex environments for robotic arms. By further restricting interval sampling to reduce poor-quality nodes, and by designing a cost function to calculate the cost of multiple candidate points, the optimal node is selected as the sampling node, thus solving the problem of low-quality path generation in complex environments for robotic arms.

[0008] The technical solution of this invention is implemented as follows: In one aspect, this invention provides a path planning method suitable for rapid convergence in complex environments for robotic arms, comprising the following steps: Step S1, Construct the starting point search tree T1: Create a set of starting tree nodes V1, containing only the starting position x of the robotic arm path. start Initialize the edge set E1 to be an empty set, and form the starting point search tree T1 by V1 and E1; Constructing the target point search tree T2: Create a target tree node set V2, containing only the target position x of the robotic arm path. goal Initialize the edge set E2 to be empty, and form the target point search tree T2 by V2 and E2; Step S2: Perform n samplings within the restricted sampling interval, filter the n sampled points in the non-obstacle interval to obtain candidate points, and then select the optimal sampling point from the candidate points; Step S3: In the initial tree node set V1, find the optimal sampling point x. rand nearest node x nearest From the nearest node xnearest To the optimal sampling point x rand The new node x is obtained by moving the direction and step size StepSize. new ; If a new node x is detected new With the nearest node x nearest If there is no collision on the paths between them, then the new node x will be... new Add to the initial tree node set V1, and in the nearest node x nearest With new node x new Add an edge between them to E1; if the new node x new With the optimal sampling point x rand If the positions overlap, mark the starting tree expansion result result1 = Reached; if they do not overlap, mark result1 = Advanced. If a collision is detected, mark the starting tree expansion result result1 = Trapped, and the starting tree will not be expanded this time; Step S4: When the initial tree expansion result result1 ≠ Trapped, find the optimal sampling point x in the target tree node set V2. rand nearest node x nearest From the nearest node x nearest To the optimal sampling point x rand The new node x is obtained by moving the direction and step size StepSize. new ; If a new node x is detected new With the nearest node x nearest The paths between them are collision-free, so the new node x will be... new Add to the target tree node set V2, and in the nearest node x nearest With new node x new Add an edge between them to E2; if the new node x new With the optimal sampling point x rand If the positions overlap, mark the starting tree expansion result result2=Reached; if they do not overlap, mark result2=Advanced. Step S5: If the target tree expansion result result2 ≠Reached, swap the roles of the starting tree and the target tree, and then proceed to the next iteration until the maximum number of iterations is reached; If the result of expanding the target tree is result2 = Reached, merge the node sets of the starting tree and the target tree into V = V1∪V2, and merge the edge sets into E = E1∪E2, forming a complete collision-free path tree T = (V, E). Then return the path tree directly to end the algorithm.

[0009] Preferably, in step S2, the sphere generated with the line connecting the current expansion point and the target point as its diameter is the restricted sampling interval.

[0010] Preferably, in step S2, the method for selecting the optimal sampling point is as follows: a and b are the most recently added nodes in the two search trees. A restricted sampling interval is performed to obtain n candidate points p. D1 represents the shortest distance from point p to the line connecting a and b, and D2 represents the distance from point p to the end point of a non-current expanded tree. Cost calculations are performed on each candidate point, and the cost function is as follows:

[0011] in, yes The weight represents the degree to which the shortest distance from point p to the line connecting point p and ab affects the cost; yes The weights represent the degree to which the distance from point b to candidate point p affects the cost; The candidate point with the lowest cost is selected as the optimal sampling point according to the above formula.

[0012] Preferably, based on the analysis of the path expansion and generation process, the weights that need to be adjusted are: Convergence Phase: Two trees are close but not yet successfully connected. The goal at this stage is to quickly connect the two trees, prioritizing the point closer to the non-current expanded tree. It should be reduced. It should be enlarged to show that more emphasis is placed on the effect of being close to another tree; Exploration Phase: The two trees are far apart in the early stages of expansion. The goal at this time is to move the route as close as possible to the line connecting the current endpoints of the two trees and advance towards the target connection. It should be increased. The value should be reduced, indicating a greater emphasis on approaching and converging towards the midline; Detour Phase: In this phase, the local environment is cluttered with obstacles or has experienced numerous expansion failures. The requirement is to allow deviations from the current endpoint of the dual-tree connection, reducing the focus on the target and expanding the search area. It should be increased. The number should be maintained or slightly reduced, reflecting a reduction in constraints on sampling points and a restoration of extensive exploration; Poor path quality: The current path is too long or not smooth. In this case, the requirement is to try connecting points closer to the current endpoint of the binary tree. It should be increased. It should be increased slightly, demonstrating an optimized path.

[0013] Preferably, during collision detection, the links of the robotic arm are abstracted as cylinders; all obstacles are expanded outward to a radius equal to the radius of the cylinders abstracted from the robotic arm links, which is the obstacle safety zone, to avoid potential collisions between each link of the robotic arm and each obstacle in the workspace before path generation.

[0014] In another aspect, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method.

[0015] In another aspect, the present invention also proposes a computer program product, comprising a computer program that, when executed by a processor, implements the above-described method.

[0016] The beneficial effects of this invention are as follows: (1) A restricted region sampling method was introduced, and the cost of candidate points was calculated. The optimal number of candidate points was defined, and the weights were adaptively adjusted in the cost function. In three-dimensional space, sampling within a spherical space with the direct line connecting the two starting points as the diameter is more likely to yield higher quality sampling points than random sampling. Calculating the cost of multiple candidate points and selecting the optimal point can more effectively reduce the randomness of the original algorithm and solve the problem of low quality generated paths. (2) Considering the requirements of high stability and security in the application environment, this paper introduces the obstacle expansion method. Before collision detection, the obstacle is expanded to a certain width, and then collision detection is performed on the expanded obstacle, which can effectively avoid unnecessary collisions caused by the generated path being too close to the obstacle; (3) To meet the needs of the robotic arm as the algorithm execution body, the algorithm is transferred from two-dimensional space to three-dimensional space for path planning, so that it can support more complex scenarios, achieve more accurate obstacle avoidance, and effectively improve the practicality of the algorithm in the actual working space of the robotic arm. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a comparison chart showing the limitation of the sampling range in this invention with existing ones.

[0019] Figure 2 This is a schematic diagram illustrating the optimization of sampling points.

[0020] Figure 3 This is a schematic diagram of the expansion of an obstacle.

[0021] Figure 4 This is a schematic diagram of expansion collision detection between a robotic arm and an obstacle. Detailed Implementation

[0022] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0023] In one aspect, this invention proposes a path planning method suitable for rapid convergence in complex environments for robotic arms, comprising the following steps: Step S1: Initialize the dual search tree and key parameters Constructing the starting point search tree T1: Create a set of starting tree nodes V1, containing only the starting position x of the robotic arm path. start Initialize the edge set E1 to be an empty set, and form the starting point search tree T1 by V1 and E1; Constructing the target point search tree T2: Create a target tree node set V2, containing only the target position x of the robotic arm path. goal Initialize the edge set E2 to be empty, and form the target point search tree T2 by V2 and E2; Initialize fusion flag: Set the dual-tree fusion flag Ef = 0, which is used to determine whether the sampling strategy needs to be adjusted later. The initial state is unfused.

[0024] Step S2, Iterative Sampling and Node Optimization

[0025] The loop continues until the maximum number of iterations, niter. For each iteration (from i=1 to i=niter), the following sampling and optimization operations are performed: Determine the sampling probability: Based on the dual-tree fusion flag Ef, calculate the sampling probability pth for the current iteration using the SampleProbability(Ef) function. This probability affects the generation direction of subsequent sampling points; before fusion, it emphasizes global exploration, while near fusion, it emphasizes precise local sampling. pth is a heuristic probability function value, which is affected by the number of failed extended collision detections. The specific calculation formula is as follows:

[0026] Here, f represents the number of collision detection failures, and its value is a non-negative integer. As f increases, pth decreases rapidly. The following example illustrates the dynamic change of pth: When an expansion failure occurs, i.e., f equals 1, the probability of sampling in the restricted area decreases, which allows the algorithm to attempt to successfully expand the node using global random sampling.

[0027] Generate constrained optimization sampling points: Combining the sampling probability pth, the preset step size StepSize, and the tree node set, candidate sampling points x are generated using the OptimizedConstrainedSampling function. candidate This step takes into account the motion constraints of the robotic arm joints (such as the range of joint angles and the limit of movement speed) to avoid generating sampling points that are outside the physical feasible domain.

[0028] Optimize sampling point location: through SamplingPointOptimization(x candidate The function fine-tunes the candidate sampling point xcandidate, further reducing the randomness of the sampling point, increasing its probability of moving closer to the target tree, and shortening the connection distance between the two trees.

[0029] In this sampling step, existing studies categorize sampling methods into global random sampling and restricted interval sampling, also known as near sampling. Near sampling generates two circles (spheres in 3D) between the current expansion point and the target point, with the line connecting them as the radius. The intersection of these two circles (spheres) is the restricted interval for near sampling. Figure 1 As shown in (a). The further restricted sampling interval in this method is defined as generating a sphere with the line connecting the current expansion point and the target point as its diameter. For example... Figure 1 As shown in (b), the process of sampling before restricting the region is optimized to sampling directly within the restricted interval. Compared with the existing algorithms' random sampling across the entire workspace and restricted interval sampling, this method further restricts the sampling interval to reduce the generation of invalid sampling points, thereby reducing computation time. In addition, this method also realizes the logical reproduction of the original sampling method in three-dimensional space, so the circle in two-dimensional space should be changed to a sphere in three-dimensional space.

[0030] In the sampling point optimization step, n samplings are performed within the restricted sampling interval. The n sampled points are then filtered for candidate points through non-obstacle interval selection. The cost of each candidate point is then calculated, and the candidate point with the lowest cost is selected as the optimal sampling point x for the restricted area. rand Return. (e.g.) Figure 2As shown, red points a and b are the most recently added nodes in random trees TreeA and TreeB, respectively. Green points represent candidate points obtained by performing one restricted interval sampling function, with a quantity of n. Blue point p is the best candidate point and the final selected sampling point. D1 represents the shortest distance from point p to the line connecting a and b; a shorter value makes it easier to generate a smooth path. D2 represents the distance from point p to a point other than the current end point of the expanded tree; a shorter value makes it easier to connect the two trees. The specific cost function is...

[0031] in, yes The weights represent the degree to which the shortest distance from point p to the line connecting point p and ab affects the cost; yes The weights represent the degree to which the distance from point b to candidate point p affects the cost; A larger value indicates that the point is further away from the line connecting the target point and the center. The path generated through this point will be more tortuous and the speed at which it approaches the target point will be slower. Conversely, a smaller value indicates that the path will be smoother and the path will approach the target point faster.

[0032] Based on the analysis of the path expansion and generation process, the scenarios requiring weight adjustment, the specific adjustments made, and the possible effects are analyzed as follows: (1) Convergence phase: The two trees are close but have not yet been successfully connected. The requirement at this time is to quickly connect the two trees, prioritizing the point closer to the non-current expanded tree. It should be reduced. It should be enlarged to show that more emphasis is placed on the effect of being close to another tree; (2) Exploration phase: The two trees are far apart in the early stage of expansion. At this time, the requirement is to move the route as close as possible to the line connecting the current ends of the two trees and advance towards the connection target. It should be increased. The value should be reduced, indicating a greater emphasis on approaching and converging towards the midline; (3) Detour phase: The current local environment is densely packed with obstacles or has experienced many failed expansion attempts. At this point, the requirement is to allow deviations from the line connecting the current endpoints of the two trees, reducing the target focus and expanding the search range. It should be increased. The number should be maintained or slightly reduced, reflecting a reduction in constraints on sampling points and a restoration of extensive exploration; (4) Poor path quality: The current path is too long or not smooth. In this case, the requirement is to try to connect the path closer to the current end point of the twin tree. It should be increased. It should be increased slightly, demonstrating an optimized path.

[0033] Step S3, Expansion and Collision Detection of the Starting Tree T1

[0034] Find the nearest node: In the initial tree node set V1, use the GetNearest(V1, xrand) function to find the node closest to the optimal sampling point x. rand The nearest node x nearest (Distance is usually calculated as Euclidean distance or Manhattan distance in the joint space of the robotic arm.) Generate new nodes: based on the preset step size StepSize and the nearest node x nearest and the optimal sampling point x rand A new node x is generated using the GetNewState function. new From the nearest node x nearest To the optimal sampling point x rand The new node x is obtained by moving the direction and step size StepSize. new ; Collision detection and tree expansion judgment: If CollisionFree(x) is used... new , x nearest The function detected a new node x. new With the nearest node x nearest If the path between them is collision-free (no obstacles and conforms to the robot arm's own motion constraints), then call Extend(V1, x rand The function will create a new node x. new Add to the initial tree node set V1, and in the nearest node x nearest With new node x new Add an edge between them to E1. If the new node x new With the optimal sampling point x rand If the positions overlap (i.e., expand to the sampling point in one step), mark the starting tree expansion result result1 = Reached; if they do not overlap (expand by one step), mark result1 = Advanced; if a collision is detected, mark the starting tree expansion result result1 = Trapped, and the starting tree will not be expanded in this case.

[0035] Step S4: Expansion of target tree T2 and determination of dual-tree fusion.

[0036] Target tree expansion trigger condition: The expansion process of target tree T2 is only started when the expansion result of starting tree T1 is not equal to Trapped (i.e., the starting tree is successfully expanded).

[0037] Similarly, in the target tree node set V2, find the optimal sampling point x. rand The nearest node x nearest From the nearest node x nearest To the optimal sampling point x rand The new node x is obtained by moving the direction and step size StepSize.new If a new node x is detected new With the nearest node x nearest The paths between them are collision-free, so the new node x will be... new Add to the target tree node set V2, and in the nearest node x nearest With new node x new Add an edge between them to E2; if the new node x new With the optimal sampling point x rand If the positions overlap, mark the starting tree expansion result result2=Reached; if they do not overlap, mark result2=Advanced.

[0038] Step S5, continue the dual-tree role swap and iteration.

[0039] If the target tree expansion result result2 ≠Reached (the two trees are not merged), swap the roles of the starting tree and the target tree (i.e., the next iteration starts with the original target tree and ends with the original starting tree) to ensure that the two trees can explore the space bidirectionally and uniformly, thus improving the fusion efficiency; then proceed to the next iteration until the maximum number of iterations niter is reached. If the target tree expansion result result2 = Reached (i.e., the target tree has been successfully expanded to the vicinity of the sampling point, and the two trees are connected), merge the node sets of the starting tree and the target tree into V = V1∪V2, and merge the edge sets into E = E1∪E2, forming a complete collision-free path tree T = (V, E), and directly return to the path tree to end the algorithm.

[0040] In this invention, the robotic arm itself has a multi-link mechanical structure, and each link has a different structural design, making the computational complexity of directly performing collision detection on the robotic arm structure extremely high. To reduce computational complexity while safely executing the algorithm for path planning, this invention abstracts the links of the robotic arm as cylinders. During path planning, all obstacles are expanded outward to a radius *r* equal to the radius of the cylinders abstracted from the robotic arm links; this is the obstacle safety zone. This avoids potential collisions between each link of the robotic arm and each obstacle in the workspace before path generation. Figure 3 A schematic diagram of the obstacle's expansion is shown.

[0041] The collision detection problem between the robotic arm and obstacles is transformed into a straight-line distance problem between a cylinder and a sphere or cube during calculation. Figure 4 As shown; The standard equation of any plane in space is: ax + by + cz + d = 0; in(a , b , c () is the normal vector of the plane, which determines the spatial orientation of the plane; d It is a parameter related to the distance from the plane to the origin of the coordinate system, which determines the position of the plane.

[0042] The essence of a collision is that "a point on the link coincides with a point on the surface of the obstacle," that is, "points on the link segment simultaneously fall on the plane of the obstacle." The position of any point on the link segment between point a (starting point) and point b (ending point) can be represented by (parametric equation of the link segment):

[0043] Substitute the parametric equation of the line segment into the plane equation. ax + by + cz + d = 0, solving for the result yields:

[0044] Where t represents the position parameter of a point on the link segment. If 0≤t≤1: the intersection point falls on the link segment (belonging to the link itself) and the point is on the surface of the obstacle, that is, the link collides with the obstacle. If molecule = 0: the starting point is on the plane (collision is determined directly); If the numerator is > 0 or < 0: the starting point is on one side of the plane (it is necessary to determine whether the line segment crosses the plane in conjunction with the denominator); If the denominator = 0: the link vector is perpendicular to the normal vector (the link is parallel to the plane) and there is no intersection (or they coincide completely); if the denominator > 0 or < 0: the link has a unique intersection with the plane (it needs to be determined by t whether it is on the link).

[0045] The above-mentioned treatment not only eliminates the potential collision risk of the path planned by the robotic arm's execution algorithm, but also effectively reduces the computational complexity.

[0046] In another aspect, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method.

[0047] In another aspect, the present invention also proposes a computer program product, comprising a computer program, characterized in that the computer program implements the above-described method when executed by a processor.

[0048] In particular, according to some embodiments of this disclosure, the processes described above can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by a processing device, it performs the functions defined above in the methods of some embodiments of this disclosure.

[0049] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a task data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated task data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0050] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol, such as HTTP (Hypertext Transfer Protocol), and can interconnect with digital task data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet), and end-to-end networks (e.g., ad hoc end-to-end networks), as well as any currently known or future-developed networks.

[0051] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to: determine the network connection status of the switch production line management application in response to detecting a query operation on a production collaboration document in the switch production line management application; replace the webpage entry information corresponding to the production collaboration document with target entry file information and load target webpage resource information in response to determining that the network connection status of the switch production line management application indicates an offline state, so as to display the webpage of the production collaboration document offline in the switch production line management application, wherein the target entry file information is the file information of the entry file corresponding to the webpage of the production collaboration document downloaded in advance, and the target webpage resource information is the resource information corresponding to the webpage stored locally; in response to determining that the network connection status of the switch production line management application indicates an online state and that the webpage resource information corresponding to the production collaboration document is not stored locally, download the webpage resource information of the webpage from the production line document server, wherein the webpage resource information includes an entry file and resource information; display the webpage of the production collaboration document in the switch production line management application according to the webpage resource information, and store the webpage resource information in a local database.

[0052] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including product-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0053] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A path planning method suitable for rapid convergence in complex environments for robotic arms, characterized in that, Includes the following steps: Step S1, Construct the starting point search tree T1: Create a set of starting tree nodes V1, containing only the starting position x of the robotic arm path. start Initialize the edge set E1 to be an empty set, and form the starting point search tree T1 by V1 and E1; Constructing the target point search tree T2: Create a target tree node set V2, containing only the target position x of the robotic arm path. goal Initialize the edge set E2 to be empty, and form the target point search tree T2 by V2 and E2; Step S2: Perform n samplings within the restricted sampling interval, filter the n sampled points in the non-obstacle interval to obtain candidate points, and then select the optimal sampling point from the candidate points; Step S3: In the initial tree node set V1, find the optimal sampling point x. rand nearest node x nearest , by the nearest node x nearest To the optimal sampling point x rand The new node x is obtained by moving the direction and step size StepSize. new ; If a new node x is detected new With the nearest node x nearest If there is no collision on the paths between them, then the new node x will be... new Add to the initial tree node set V1, and in the nearest node x nearest With new node x new Add an edge between them to E1; if the new node x new With the optimal sampling point x rand If the positions overlap, mark the starting tree expansion result result1 = Reached; if they do not overlap, mark result1 = Advanced. If a collision is detected, mark the starting tree expansion result result1 = Trapped, and the starting tree will not be expanded this time; Step S4: When the initial tree expansion result result1 ≠ Trapped, find the optimal sampling point x in the target tree node set V2. rand nearest node x nearest , by the nearest node x nearest To the optimal sampling point x rand The new node x is obtained by moving the direction and step size StepSize. new ; If a new node x is detected new With the nearest node x nearest The paths between them are collision-free, so the new node x will be... new Add to the target tree node set V2, and in the nearest node x nearest With new node x new Add an edge between them to E2; if the new node x new With the nearest node x rand If the positions overlap, mark the starting tree expansion result result2=Reached; if they do not overlap, mark result2=Advanced. Step S5: If the target tree expansion result result2 ≠Reached, swap the roles of the starting tree and the target tree, and then proceed to the next iteration until the maximum number of iterations is reached; If the result of expanding the target tree is result2 = Reached, merge the node sets of the starting tree and the target tree into V = V1∪V2, and merge the edge sets into E = E1∪E2, forming a complete collision-free path tree T = (V, E). Then return directly to this path tree and the algorithm ends.

2. The path planning method for rapid convergence in complex environments for robotic arms as described in claim 1, characterized in that, In step S2, the sphere generated with the line connecting the current expansion point and the target point as its diameter is the restricted sampling interval.

3. The path planning method for rapid convergence in complex environments for robotic arms as described in claim 2, characterized in that, In step S2, the method for selecting the optimal sampling point is as follows: a and b are the most recently added nodes in the two search trees. A restricted sampling interval is performed to obtain n candidate points p. D1 represents the shortest distance from point p to the line connecting a and b, and D2 represents the distance from point p to the end point of a non-current expanded tree. Cost calculations are performed on each candidate point, and the cost function is as follows: in, yes The weight represents the degree to which the shortest distance from point p to the line connecting point p and ab affects the cost; yes The weights represent the degree to which the distance from point b to candidate point p affects the cost; The candidate point with the lowest cost is selected as the optimal sampling point according to the above formula.

4. The path planning method for rapid convergence in complex environments for robotic arms as described in claim 3, characterized in that, Based on the analysis of the path expansion and generation process, the following weight adjustments are required: Convergence Phase: Two trees are close but not yet successfully connected. The goal at this stage is to quickly connect the two trees, prioritizing the point closer to the non-current expanded tree. It should be reduced. It should be enlarged to show that more emphasis is placed on the effect of being close to another tree; Exploration Phase: The two trees are far apart in the early stages of expansion. The goal at this point is to move the route as close as possible to the line connecting the current endpoints of the two trees and advance towards the target connection. It should be increased. The value should be reduced, indicating a greater emphasis on approaching and converging towards the midline; Detour Phase: In this phase, the local environment is cluttered with obstacles or has experienced numerous expansion failures. The requirement is to allow deviations from the current endpoint of the dual-tree connection, reducing the focus on the target and expanding the search area. It should be increased. The number should be maintained or slightly reduced, reflecting a reduction in constraints on sampling points and a restoration of extensive exploration; Poor path quality: The current path is too long or not smooth. In this case, the requirement is to try connecting points closer to the current endpoint of the binary tree. It should be increased. It should be increased slightly, demonstrating an optimized path.

5. The path planning method for rapid convergence in complex environments for robotic arms as described in claim 1, characterized in that, During collision detection, the links of the robotic arm are abstracted as cylinders; all obstacles are expanded outward to a radius equal to the radius of the cylinders abstracted from the robotic arm links, which is the obstacle safety zone. This avoids potential collisions between each link of the robotic arm and each obstacle in the workspace before path generation.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 5.

7. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 5.

Images