A path planning method for collaborative control of overhead crane mobile robotic arms

Abstract

This invention discloses a path planning method for collaborative control of a crane and a mobile robotic arm, relating to the field of mobile robotic arm path planning. The method includes: obtaining the position information of the mobile robotic arm and the crane based on a two-dimensional plan view of the factory; combining the position information with a path planning algorithm to generate a feasible path and calculating the path length of the feasible path; establishing a hyperelliptical intelligent constraint sampling condition with the path length of the feasible path as the major axis, and generating elliptical constraints based on sampling points obtained from the hyperelliptical intelligent constraint sampling condition; combining the elliptical constraints with a bidirectional growth direct connection strategy to generate a direct connection path and calculating the path length of the direct connection path; obtaining the optimal planned path based on the path length of the direct connection path and an optimization strategy; and generating drive commands based on the optimal planned path to control the movement path of the mobile robotic arm and the crane. This invention can avoid situations where the crane's movement speed exceeds the robotic arm's following capability due to insufficient path planning speed.

Description

Technical Field

[0001] This invention relates to the field of path planning for mobile robotic arms, and more specifically, to a path planning method for the collaborative control of a crane mobile robotic arm. Background Technology

[0002] In recent years, path planning research has continuously introduced new theories and technologies, which is particularly evident in the RRT series of path planning algorithms. In order to improve the performance of the RRT series algorithms, existing technologies have optimized the path planning algorithms in terms of path length, time cost, search efficiency and robustness by introducing new theories or technologies.

[0003] Some approaches optimize path search strategies. For example, the existing technology proposes a three-source fast expanding random tree GT-RRT algorithm that combines a gravitational field to guide the generation of new nodes, greatly reducing the search path length. Therefore, it is believed that research on probability-based algorithms for improving the robustness of path planning is still relatively rare. Current research on path planning mainly focuses on optimizing path length, time cost, and search efficiency, as well as integrating different algorithms according to specific needs. This approach can better adapt to the needs of different scenarios and problems, and improve the efficiency and accuracy of path planning.

[0004] For example, in the field of path planning technology for overhead crane mobile robotic arms, path planning is widely used. However, in the existing technology, the path search efficiency of the mobile robotic arm is low when facing complex road conditions, which leads to the inability of the overhead crane and the robotic arm to work together effectively. In other words, the low path planning efficiency causes the overhead crane to move before the mobile robotic arm.

[0005] There are currently no effective solutions to the problems in the relevant technologies. Summary of the Invention

[0006] (a) Technical problems to be solved

[0007] To address the shortcomings of existing technologies, this invention provides a path planning method for the collaborative control of a crane and a mobile robotic arm. This method enables collaborative operation between the robotic arm and the crane during heavy load transportation, avoiding situations where the crane's movement speed exceeds the robotic arm's following ability due to insufficient path planning speed. This achieves the goal of solving the problem of low path planning efficiency causing the crane to move before the mobile robotic arm.

[0008] (II) Technical Solution

[0009] To achieve the above-mentioned goal of ensuring coordinated operation between the robotic arm and the overhead crane during heavy object transportation, and avoiding the overhead crane's movement speed exceeding the robotic arm's following capability due to insufficient path planning speed, the specific technical solution adopted by this invention is as follows:

[0010] A path planning method for collaborative control of a crane's mobile robotic arm, the path planning method comprising:

[0011] S1. Obtain the position information of the mobile robotic arm and the overhead crane based on the two-dimensional plan of the factory, combine the position information with the path planning algorithm to generate a feasible path, and calculate the path length of the feasible path;

[0012] S2. Establish hyperelliptical intelligent constraint sampling conditions with the path length of feasible paths as the major axis, and generate elliptical constraints by obtaining sampling points based on the hyperelliptical intelligent constraint sampling conditions.

[0013] S3. Combine the elliptical constraint with the bidirectional growth direct connection strategy to generate a direct connection path, and calculate the path length of the direct connection path. Obtain the optimal planning path based on the path length of the direct connection path and the optimization strategy.

[0014] S4. Generate driving instructions based on the optimal planning path to control the movement path of the mobile robotic arm and the overhead crane, and introduce a dynamic environment monitoring mechanism to monitor and adjust the movement status of the mobile robotic arm and the overhead crane in real time.

[0015] Preferably, the location information of the mobile robotic arm and the overhead crane is obtained based on the two-dimensional plan of the factory. The location information is combined with the path planning algorithm to generate a feasible path, and the path length of the feasible path is calculated, including:

[0016] S11. Extract the two-dimensional plan of the factory to obtain the initial position and target position of the mobile robotic arm and the overhead crane, as well as the location information of obstacles inside the factory;

[0017] S12. Set the initial position of the mobile robotic arm as the starting position and the target position as the ending position, and use the path planning algorithm to generate an initial search tree, defining the starting position as the root point of the search tree;

[0018] S13. Generate random sampling points based on the starting and ending positions, find nodes in the search tree whose distance from the random sampling points is a threshold, and generate new nodes by extending them by a preset length in the direction from the nodes to the random sampling points.

[0019] S14. Based on the obstacle location information, determine whether there is a collision between the new node and the node. If a collision occurs, discard the path. If no collision occurs, add the new node to the search tree and connect the random sampling point with the new node.

[0020] S15. Determine the distance relationship between the new node and the endpoint, generate a feasible path based on the distance result, and record the path length of the feasible path.

[0021] Preferably, establishing hyperelliptical intelligent constraint sampling conditions with the path length of feasible paths as the major axis, and generating elliptical constraints based on sampling points obtained from the hyperelliptical intelligent constraint sampling conditions, includes:

[0022] S21. Define a superellipse based on superellipse intelligent constraint sampling, and randomly sample points within the unit circle according to the sampling rules, then transfer the random points to the superellipse.

[0023] S22. Determine the center point of the hyperellipse based on the starting and ending positions of the mobile robotic arm on the horizontal and vertical axes, and define the hyperellipse radius information using the path length of the feasible path and the straight-line distance from the starting and ending positions of the mobile robotic arm. Determine the position of the hyperellipse based on the radius information and the center.

[0024] S23. Use the hyperellipse position to determine the hyperellipse constraint sampling parameters, and determine the sampling points inside the hyperellipse based on the hyperellipse constraint sampling parameters.

[0025] Preferably, the elliptical constraint is combined with a bidirectional growth direct connection strategy to generate a direct connection path, and the path length of the direct connection path is calculated. The optimal planned path is obtained based on the path length of the direct connection path and the optimization strategy, including:

[0026] S31. Construct a forward growth tree and a reverse growth tree with the starting position and ending position of the mobile robotic arm as the root points respectively, and generate forward random sampling points and reverse random sampling points based on elliptical constraints using the forward growth tree and the reverse growth tree.

[0027] S32. In the forward growth tree and the reverse growth tree, find the forward node and the reverse node whose distance from the forward random sampling point and the reverse random sampling point is a threshold, respectively.

[0028] S33. Extend the forward node to the forward random sampling point by a preset length, and extend the reverse node to the reverse random sampling point by a preset length, and generate a new forward node and a new reverse node based on the extension results.

[0029] S34. Generate a direct path based on the new forward node and the new reverse node. After determining the existence of obstacles, calculate the path length of the direct path and obtain the optimal planned path by combining the optimization strategy.

[0030] Preferably, a direct path is generated based on the new forward node and the new reverse node. After determining the existence of obstacles, the path length of the direct path is calculated, and the optimal planned path is obtained by combining the optimization strategy, including:

[0031] S341. Connect the new forward node and the forward random sampling point to obtain the forward search path, and connect the new reverse node and the reverse random sampling point to obtain the reverse search path. Determine the existence of obstacles on the forward search path and the reverse search path respectively.

[0032] S342. If there are obstacles, discard the corresponding path. If there are no obstacles, add the new forward node and the new reverse node to the forward growth tree and the reverse growth tree respectively, and connect the new forward node and the new reverse node to obtain a direct path.

[0033] S343. Calculate the path lengths of the forward search path, the reverse search path, and the direct connection path respectively. At the same time, determine the traversal optimization parameters based on the optimization strategy, and obtain the optimal planned path based on the traversal optimization parameters and the path length.

[0034] The preferred formula for calculating the path length of the forward search path is:

[0035]

[0036] The formula for calculating the path length of the reverse search path is:

[0037]

[0038] The formula for calculating the path length of a directly connected path is:

[0039] L z =d(P znew ,P fnew );

[0040] In the formula, L1 represents the path length of the forward search path, L2 represents the path length of the reverse search path, and L... z The path length of the direct connection is represented by n, the dimension is represented by d, and the Euclidean distance is represented by P. zi P represents a forward node. zi+1 P represents the (i+1)th forward node. fi P represents the reverse node. fi+1 P represents the (i+1)th reverse node. znew P represents a new node in the forward direction. fnew This represents a new node in reverse order, and i represents the number of nodes.

[0041] Preferably, the method generates drive commands based on the optimal planned path to control the movement path of the mobile robotic arm and the overhead crane, and introduces a dynamic environment monitoring mechanism to monitor and adjust the movement status of the mobile robotic arm and the overhead crane in real time, including:

[0042] S41. Generate driving instructions based on the optimal planning path and send the driving instructions to the mobile robotic arm drive controller. The mobile robotic arm and the overhead crane move according to the optimal planning path.

[0043] S42. Using sensors and radar within the factory, monitor the movement of the mobile robotic arm, as well as information about the factory environment and obstacles, and transmit the monitoring results to the central processing unit.

[0044] S43. The central processing unit receives the monitoring results, identifies abnormal states, generates decision instructions based on the abnormal states to control the mobile robotic arm, and regenerates the planned path to control the movement state of the mobile robotic arm.

[0045] Preferably, the process involves generating drive commands based on the optimal planned path and sending these commands to the mobile robotic arm drive controller. The mobile robotic arm and the overhead crane then move according to the optimal planned path, including:

[0046] S411. Generate path points based on the optimal planning path, analyze the joint angles of each joint of the mobile robotic arm when executing the path points, and determine the target speed based on the joint angles and the initial joint angles of the mobile robotic arm.

[0047] S412. Combine the joint angle with the target speed as the driving command for the mobile robotic arm, and send the driving command to the mobile robotic arm drive controller through the communication interface.

[0048] S413 The mobile robotic arm drive controller receives drive commands to control the movement of the mobile robotic arm. During the movement, the overhead crane follows the mobile robotic arm.

[0049] Preferably, the process involves generating path points based on the optimal planned path, analyzing the joint angles of each joint of the mobile robotic arm when executing path points, and determining the target speed based on the joint angles and the initial joint angles of the mobile robotic arm, including:

[0050] S4111. Generate a grid map based on the factory's two-dimensional plan view, and generate path Voronoi points on the grid map according to the optimal planning path and Voronoi map generation rules.

[0051] S4112. Traverse the path through the Voronoi points, select the intersection points, and connect adjacent intersections. Extract guide points based on the relationship between the lines seen at the intersection points and obstacles, and use the guide points as path points.

[0052] S4113. Use computer simulation technology to establish a finite element model of the mobile robotic arm, and simulate the target joint angle of the mobile robotic arm when it moves along the path point in the working process according to inverse kinematics within the finite element model.

[0053] S4114. Obtain the initial joint angles of the mobile robotic arm, and determine the joint angular velocity of the mobile robotic arm based on the initial joint angles and the target joint angles.

[0054] Preferably, the formula for calculating the joint angular velocity of the robotic arm is:

[0055] T = k p (T target -T cur );

[0056] In the formula, T represents the joint angular velocity of the moving robotic arm, and k p T represents the proportional gain. target T represents the target joint angle. cur This indicates the initial joint angle.

[0057] (III) Beneficial Effects

[0058] Compared with existing technologies, this invention provides a path planning method for the collaborative control of a crane's mobile robotic arm, which has the following advantages:

[0059] (1) This invention optimizes the path planning algorithm of the mobile robotic arm to achieve efficient path search, which can effectively improve the crane's ability to follow the mobile robotic arm and ensure the coordinated operation between the robotic arm and the crane during the transportation of heavy objects. It avoids the crane's movement speed exceeding the robotic arm's following ability due to insufficient path planning speed, thereby reducing transportation time and improving overall operation efficiency.

[0060] (2) The path planning method for collaborative control of overhead crane and mobile robotic arm proposed in this invention has the advantage of low time cost in searching for the first feasible path, so as to obtain a feasible path at the fastest speed and calculate its path length as an undetermined variable of hyperelliptic constraint, thereby quickly establishing sampling constraints, which can optimize the tracking ability of mobile robotic arm and thus solve the problem of collaborative operation. Attached Figure Description

[0061] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments 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.

[0062] Figure 1 This is a flowchart of a path planning method for collaborative control of a crane-moving robotic arm according to an embodiment of the present invention;

[0063] Figure 2 This is a bidirectional growth defect diagram according to an embodiment of the present invention;

[0064] Figure 3 This is an elliptical constraint defect diagram according to an embodiment of the present invention;

[0065] Figure 4 This is an algorithm flowchart according to an embodiment of the present invention;

[0066] Figure 5 The IBTO-RRT algorithm is used to first reconnoiter the map according to an embodiment of the present invention;

[0067] Figure 6 This is a bidirectional search graph of the IBTO-RRT algorithm according to an embodiment of the present invention;

[0068] Figure 7 This is a bidirectional direct connection diagram of the IBTO-RRT algorithm according to an embodiment of the present invention;

[0069] Figure 8 This is a sampling constraint diagram of the IBTO-RRT algorithm according to an embodiment of the present invention;

[0070] Figure 9 This is a path planning simulation diagram under a simple RRT environment according to an embodiment of the present invention;

[0071] Figure 10 This is a path planning simulation diagram under a simple RRT* environment according to an embodiment of the present invention;

[0072] Figure 11 This is a path planning simulation diagram under a simple Informed-RRT* environment according to an embodiment of the present invention;

[0073] Figure 12 This is a path planning simulation diagram under a simple environment based on the BITO-RRT* embodiment of the present invention;

[0074] Figure 13 This is a path planning simulation diagram under a simple environment according to an embodiment of the present invention (IBTO-RRT*).

[0075] Figure 14 This is a path planning simulation diagram under a complex RRT environment according to an embodiment of the present invention;

[0076] Figure 15 This is a simulation diagram of path planning in a complex environment under RRT* according to an embodiment of the present invention;

[0077] Figure 16 This is a simulation diagram of path planning in a complex environment according to an embodiment of the present invention using Informed-RRT*.

[0078] Figure 17 This is a path planning simulation diagram under complex environment according to an embodiment of the present invention;

[0079] Figure 18 This is a path planning simulation diagram under complex environment according to an embodiment of the present invention;

[0080] Figure 19 This is a path length comparison diagram according to an embodiment of the present invention;

[0081] Figure 20 This is a comparison chart of time costs according to an embodiment of the present invention;

[0082] Figure 21 This is a comparison chart of comprehensive performance according to an embodiment of the present invention. Detailed Implementation

[0083] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention.

[0084] According to an embodiment of the present invention, a path planning method for collaborative control of a crane's mobile robotic arm is provided.

[0085] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 As shown, the path planning method for cooperative control of a crane-mounted mobile robotic arm according to an embodiment of the present invention includes:

[0086] S1. Obtain the position information of the mobile robotic arm and the overhead crane based on the two-dimensional plan of the factory, combine the position information with the path planning algorithm to generate a feasible path, and calculate the path length of the feasible path.

[0087] In this embodiment, the position information of the mobile robotic arm and the overhead crane is obtained based on the two-dimensional plan of the factory. The position information is combined with the path planning algorithm to generate a feasible path, and the path length of the feasible path is calculated, including:

[0088] S11. Extract the two-dimensional plan of the factory to obtain the initial position and target position of the mobile robotic arm and the overhead crane, as well as the location information of obstacles inside the factory;

[0089] S12. Set the initial position of the mobile robotic arm as the starting position and the target position as the ending position, and use the path planning algorithm to generate an initial search tree, defining the starting position as the root point of the search tree;

[0090] S13. Generate random sampling points based on the starting and ending positions, find nodes in the search tree whose distance from the random sampling points is a threshold, and generate new nodes by extending them by a preset length in the direction from the nodes to the random sampling points.

[0091] S14. Based on the obstacle location information, determine whether there is a collision between the new node and the node. If a collision occurs, discard the path. If no collision occurs, add the new node to the search tree and connect the random sampling point with the new node.

[0092] S15. Determine the distance relationship between the new node and the endpoint, generate a feasible path based on the distance result, and record the path length of the feasible path.

[0093] S2. Establish hyperelliptical intelligent constraint sampling conditions with the path length of feasible paths as the major axis, and generate elliptical constraints by obtaining sampling points based on the hyperelliptical intelligent constraint sampling conditions.

[0094] In this embodiment, establishing a hyperelliptical intelligent constraint sampling condition with the path length of the feasible path as the major axis, and generating elliptical constraints based on the sampling points obtained from the hyperelliptical intelligent constraint sampling condition includes:

[0095] S21. Define a superellipse based on superellipse intelligent constraint sampling, and randomly sample points within the unit circle according to the sampling rules, then transfer the random points to the superellipse.

[0096] S22. Determine the center point of the hyperellipse based on the starting and ending positions of the mobile robotic arm on the horizontal and vertical axes, and define the hyperellipse radius information using the path length of the feasible path and the straight-line distance from the starting and ending positions of the mobile robotic arm. Determine the position of the hyperellipse based on the radius information and the center.

[0097] S23. Use the hyperellipse position to determine the hyperellipse constraint sampling parameters, and determine the sampling points inside the hyperellipse based on the hyperellipse constraint sampling parameters.

[0098] S3. Combine the elliptical constraint with the bidirectional growth direct connection strategy to generate a direct connection path, calculate the path length of the direct connection path, and obtain the optimal planning path based on the path length of the direct connection path and the optimization strategy.

[0099] In this embodiment, the elliptical constraint is combined with the bidirectional growth direct connection strategy to generate a direct connection path, and the path length of the direct connection path is calculated. The optimal planned path is obtained based on the path length of the direct connection path and the optimization strategy, including:

[0100] S31. Construct a forward growth tree and a reverse growth tree with the starting position and ending position of the mobile robotic arm as the root points respectively, and generate forward random sampling points and reverse random sampling points based on elliptical constraints using the forward growth tree and the reverse growth tree.

[0101] S32. In the forward growth tree and the reverse growth tree, find the forward node and the reverse node whose distance from the forward random sampling point and the reverse random sampling point is a threshold, respectively.

[0102] S33. Extend the forward node to the forward random sampling point by a preset length, and extend the reverse node to the reverse random sampling point by a preset length, and generate a new forward node and a new reverse node based on the extension results.

[0103] S34. Generate a direct path based on the new forward node and the new reverse node. After determining the existence of obstacles, calculate the path length of the direct path and obtain the optimal planned path by combining the optimization strategy.

[0104] Specifically, based on the generation of direct paths between new forward and reverse nodes, the path length of the direct path is calculated after determining the existence of obstacles, and the optimal planned path is obtained by combining the optimization strategy, including:

[0105] S341. Connect the new forward node and the forward random sampling point to obtain the forward search path, and connect the new reverse node and the reverse random sampling point to obtain the reverse search path. Determine the existence of obstacles on the forward search path and the reverse search path respectively.

[0106] S342. If there are obstacles, discard the corresponding path. If there are no obstacles, add the new forward node and the new reverse node to the forward growth tree and the reverse growth tree respectively, and connect the new forward node and the new reverse node to obtain a direct path.

[0107] S343. Calculate the path lengths of the forward search path, the reverse search path, and the direct connection path respectively. At the same time, determine the traversal optimization parameters based on the optimization strategy, and obtain the optimal planned path based on the traversal optimization parameters and the path length.

[0108] S4. Generate driving instructions based on the optimal planning path to control the movement path of the mobile robotic arm and the overhead crane, and introduce a dynamic environment monitoring mechanism to monitor and adjust the movement status of the mobile robotic arm and the overhead crane in real time.

[0109] In this embodiment, driving commands are generated based on the optimal planned path to control the movement paths of the mobile robotic arm and the overhead crane. A dynamic environment monitoring mechanism is introduced to monitor and adjust the movement status of the mobile robotic arm and the overhead crane in real time, including:

[0110] S41. Generate driving instructions based on the optimal planning path and send the driving instructions to the mobile robotic arm drive controller. The mobile robotic arm and the overhead crane move according to the optimal planning path.

[0111] S42. Using sensors and radar within the factory, monitor the movement of the mobile robotic arm, as well as information about the factory environment and obstacles, and transmit the monitoring results to the central processing unit.

[0112] S43. The central processing unit receives the monitoring results, identifies abnormal states, generates decision instructions based on the abnormal states to control the mobile robotic arm, and regenerates the planned path to control the movement state of the mobile robotic arm.

[0113] Specifically, driving commands are generated based on the optimal planned path and sent to the mobile robotic arm drive controller. The mobile robotic arm and the overhead crane move according to the optimal planned path, including:

[0114] S411. Generate path points based on the optimal planning path, analyze the joint angles of each joint of the mobile robotic arm when executing the path points, and determine the target speed based on the joint angles and the initial joint angles of the mobile robotic arm.

[0115] S412. Combine the joint angle with the target speed as the driving command for the mobile robotic arm, and send the driving command to the mobile robotic arm drive controller through the communication interface.

[0116] S413 The mobile robotic arm drive controller receives drive commands to control the movement of the mobile robotic arm. During the movement, the overhead crane follows the mobile robotic arm.

[0117] The process includes generating path points based on the optimal planned path, analyzing the joint angles of each joint of the mobile robotic arm when executing path points, and determining the target speed based on the joint angles and the initial joint angles of the mobile robotic arm.

[0118] S4111. Generate a grid map based on the factory's two-dimensional plan view, and generate path Voronoi points on the grid map according to the optimal planning path and Voronoi map generation rules.

[0119] S4112. Traverse the path through the Voronoi points, select the intersection points, and connect adjacent intersections. Extract guide points based on the relationship between the lines seen at the intersection points and obstacles, and use the guide points as path points.

[0120] S4113. Use computer simulation technology to establish a finite element model of the mobile robotic arm, and simulate the target joint angle of the mobile robotic arm when it moves along the path point in the working process according to inverse kinematics within the finite element model.

[0121] S4114. Obtain the initial joint angles of the mobile robotic arm, and determine the joint angular velocity of the mobile robotic arm based on the initial joint angles and the target joint angles.

[0122] Specifically, the formula for calculating the joint angular velocity of the robotic arm is as follows:

[0123] T = k p (T target -T cur );

[0124] In the formula, T represents the joint angular velocity of the moving robotic arm, and k p T represents the proportional gain. target T represents the target joint angle. cur This indicates the initial joint angle.

[0125] It should be explained that the BITO-RRT* proposed in this embodiment accelerates the search speed through a bidirectional growth strategy. Although this reduces the search time, it increases the cost of finding the first feasible path. Figure 2 As shown, Informed-RRT* relies on existing feasible paths to establish hyperelliptical constraints, significantly reducing invalid sampling points and thus accelerating the search speed. However, in complex environments, the search speed for the first feasible path is similar to that of RRT, which weakens its advantage and affects the robustness of the time cost. Figure 3 As shown, in general, Informed-RRT* improves robustness to path length but reduces robustness to time cost, while BITO-RRT* improves to time cost but still needs to improve robustness to path length. Therefore, the ideal new algorithm should optimize both search time and path length while ensuring high robustness to meet the needs of different scenarios and reduce dependence on specific algorithms.

[0126] This embodiment is mainly based on the BITO-RRT* algorithm and the Informed-RRT* algorithm. It points out that the former has fewer effective nodes and lower search efficiency due to the randomness of sampling, while the latter is difficult to establish hyperelliptic constraints in complex situations. Therefore, this embodiment uses the sampling constraint method of the Informed-RRT* algorithm to improve the sampling efficiency of the BITO-RRT algorithm. At the same time, it uses the BITO-RRT* algorithm to solve the problem of high time cost of the Informed-RRT* algorithm. In addition, the RRT algorithm is used to search and establish elliptic constraints first to deal with the difficulty of establishing sampling constraints in bidirectional search. Finally, an improved algorithm IBTO-RRT* based on elliptic constraints and bidirectional search is proposed to solve the coordination problem of overhead crane and mobile robotic arm.

[0127] This embodiment proposes the IBTO-RRT* algorithm, which leverages the strengths and compensates for the weaknesses of traditional algorithms. First, it utilizes the short search time of the RRT algorithm to identify a non-optimal path as the foundation for constraint sampling (i.e., the first-look method). Then, it employs a bidirectional growth strategy with direct connection detection to accelerate convergence. Simultaneously, it introduces a sampling constraint strategy with hyperelliptic constraints to reduce the acquisition of invalid points. Next, it optimizes the direct path through a traversal optimization step. Finally, it reselects nodes in the asymptotic convergence region to shorten the path length. The advantages of the IBTO-RRT algorithm lie in its excellent performance across all aspects, significantly reducing the constraints imposed by requirements. While optimizing time costs and path search, it also increases the robustness of the path planning algorithm. Its design flowchart is as follows: Figure 4 As shown.

[0128] The IBTO-RRT algorithm mainly encompasses methods such as first-look strategy, hyperelliptic sampling constraints, bidirectional search, bidirectional direct connection, and traversal optimization. Firstly, in the RRT first-look strategy, a non-optimal path length is determined using the RRT algorithm and assigned to the major axis of the hyperellipse, thereby accelerating the formation of sampling constraints, reducing random sampling time, and lowering the algorithm's time cost. Figure 5 As shown; next, a bidirectional search is introduced, generating a forward tree and a reverse tree from the starting point and the ending point respectively, using information from both sides to search simultaneously, thereby improving efficiency, especially in complex search spaces, such as... Figure 6 As shown; the bidirectional direct connection strategy generates the path immediately upon detecting a direct connection between the start and end points, significantly reducing computational costs; if the direct connection fails, an intermediate node is generated for retry, as shown. Figure 7 As shown; the hyperelliptic sampling constraint restricts the sampling space by dynamically adjusting the major axis (current path length) and minor axis (based on the straight-line distance between the starting point and the target point), reducing invalid sampling points and accelerating path search, such as... Figure 8 As shown; finally, in the traversal optimization step, the total length of the forward, reverse and direct paths is calculated and the updated path length is recorded until the iteration is completed, and the optimal path length is finally obtained. By combining these methods, the IBTO-RRT algorithm can perform path planning efficiently.

[0129] To facilitate understanding of the above technical solutions of the present invention, the working principle or operation method of the present invention in actual process will be described in detail below.

[0130] Step 1: Extract the 2D plan of the factory to obtain the initial position, target position, and obstacle position information of the mobile robotic arm and overhead crane.

[0131] Step 2: Utilize the RRT (Path Reconnaissance and Path Response) algorithm to quickly find a non-optimal feasible path and calculate its length to accelerate the construction of sampling constraints. This is known as the first-look-back method.

[0132] (1) Determine the starting and ending positions. Take the initial position of the moving robotic arm as the starting position and the target position as the ending position. Initialize an empty search tree with the starting position of the moving robotic arm as the root point, and generate a random sampling point q in the two-dimensional plane where the starting and ending positions of the moving robotic arm exist. rand .

[0133] (2) Find the match q in the search tree rand The nearest node q neares From q nearest To q rand The direction is extended by 1 meter using Euclidean distance to obtain a new node q. new The Euclidean distance is calculated as follows:

[0134]

[0135] In the formula, x and y represent the coordinates of two points, and n represents the dimension, which is 1 here. The detection starts from q. neares to q new Check if the path collides with an obstacle. If a collision occurs, discard the extension; otherwise, set the new state q. new Add to the tree and connect q. neares and q new .

[0136] (3) After each new node is added, check whether the new node is close to the target position. If it is close, try to connect the new node to the target position and perform collision detection. Otherwise, continue to repeat step (2). If the connection to the target position is successful and there is no collision on the path, construct a path from the starting position to the ending position and record the path length. At this time, the moving robot arm is still in the starting position.

[0137] Step 3: Introduce hyperelliptical intelligent constraint sampling. After constructing sampling constraints using the path length obtained from the first-look strategy as the hyperellipse construction condition, the sampling area from the initial position of the mobile robotic arm to the target position is restricted to reduce the acquisition of invalid sampling points, thereby improving the sampling quality. Specifically:

[0138] (1) Define the hyperellipse constraint. The hyperellipse is a generalization of the ellipse, and its equation can be expressed as:

[0139]

[0140] In the formula, a and b are the radii of the hyperellipse in the x-axis and y-axis directions, respectively, and n is a positive integer used to control the shape of the hyperellipse. When n = 2, the above formula represents a standard ellipse equation.

[0141] (2) Random sampling is performed within the unit circle, and the coordinates (u, v) of the sampling points satisfy: u 2 +v 2 The values ​​are less than or equal to 1, and u and v are randomly selected from a uniform distribution. The random points within the unit circle are then transformed onto the hyperellipse. The transformation steps are as follows: x = a * u 1 / n y = b * v 1 / n The transformed (x, y) will lie on the superellipse defined by parameters a, b, and n.

[0142] (3) Determine the sampling parameters and position of the hyperellipse constraint: the major axis 2a is the length of the non-optimal feasible path detected by the first-look strategy, and the minor axis 2b is the straight-line distance C from the starting position to the ending position of the moving robot arm. minThe center point (x0, y0) of the ellipse is the starting position (x0, y0). start y start ) and endpoint position (x goal y goal The midpoint of () is calculated using the following formula:

[0143]

[0144] C best =C pioner ;

[0145]

[0146]

[0147]

[0148]

[0149]

[0150] In the formula, C pioner C represents the length of the path first detected. best This represents the current optimal path length, according to C. best As the sampling area is continuously updated, it needs to be transformed. To achieve uniform sampling within the ellipse, sampling is first performed within a unit circle, and then these points are mapped onto the actual elliptical region. This involves the application of the x-rotation angle and rotation matrix, where the rotation angle θ is:

[0151]

[0152] The rotation matrix R(θ) is:

[0153]

[0154] Final sampling point (x) final y final )for:

[0155]

[0156] In the formula, (x,y) are the points after the hyperellipse transformation. (2) to (3) are repeated until the sampling is completed.

[0157] It needs to be explained that the hyperelliptical intelligent constraint sampling is mainly divided into two parts. One part is to search for a feasible path by first-look, and the other part is elliptical constraint. Hyperelliptical constraint is used to continuously narrow the sampling area to improve the sampling quality. The length of the feasible path currently searched is used as a variable. Here, the path is obtained through the first-look strategy. Through a series of mathematical transformations, an elliptical domain is obtained. As the search path length shortens, the elliptical domain will also become smaller and smaller. Ideally, when the path length approaches the straight-line distance between the starting point and the ending point, the elliptical domain approaches a line segment from the starting point to the ending point.

[0158] Step 4: Employ bidirectional growth and bidirectional direct connection strategies to accelerate path convergence. The specific steps are as follows:

[0159] (1) Construct an empty search tree with the starting position of the mobile robotic arm as the root point as the forward growth tree, and construct an empty search tree with the ending position of the mobile robotic arm as the root point as the reverse growth tree. Generate a random sampling point q in the two trees in the two-dimensional plane containing the starting and ending positions of the mobile robotic arm. zrand q frand .

[0160] (2) Find the corresponding nodes in the two search trees. zrand q frand The nearest node q znearest With q fnearest From q znearest To q zrand and q fnearest To q frand Each direction is extended by 1 meter using Euclidean distance to obtain a new node q. znew With q fnew The Euclidean distance is calculated as follows:

[0161]

[0162] Where, x z and y z These are the coordinates of two points in the forward search tree, x f and y f These are the coordinates of two points in the reverse search tree, where n is the dimension number (here, n=1). The detection starts from q. znearest to q znew and q fnearest to q fnew Check if the path collides with an obstacle. If a collision occurs, discard the extension; otherwise, set the new state q. znew and q fnew Add to the tree and connect q. znearest and q znew and q fnearest and qfnew .

[0163] (3) Detecting the new positive sampling point q znew and the reverse new sampling point q fnew If there are no obstacles, a feasible path is generated and its length is calculated. If there are obstacles, steps (1) to (3) are repeated. The path length is calculated using the Euclidean distance formula as follows:

[0164] The formula for calculating the path length of a forward search path is:

[0165]

[0166] The formula for calculating the path length of the reverse search path is:

[0167]

[0168] The formula for calculating the path length of a directly connected path is:

[0169]

[0170] In the formula, L1 represents the path length of the forward search path, L2 represents the path length of the reverse search path, and L... z L represents the path length of a directly connected path. b Let P represent the total length of the search path, n represent the number of dimensions, d represent the Euclidean distance, and P represent the total length of the search path. zi P represents a forward node. zi+1 P represents the (i+1)th forward node. fi P represents the reverse node. fi+1 P represents the (i+1)th reverse node. znew P represents a new node in the forward direction. fnew This represents a new node in reverse order, and i represents the number of nodes.

[0171] It needs to be explained that by utilizing bidirectional growth and bidirectional direct connection strategies, the scope of path search is reduced to some extent. The bidirectional growth strategy typically constructs two exploration expansion trees: one grows from the starting point (forward exploration expansion tree) and the other grows from the ending point (reverse exploration expansion tree). These two trees expand simultaneously in each other's direction. During this process, direct connection checks are performed on the newly generated random sampling points of the two trees. If there are no obstacles between the two sampling points, a feasible path is generated, and the length of the path is calculated; otherwise, the search continues. Specifically, the path search is divided into three segments: the forward growth path, the reverse growth path, and the direct connection path. The final path length is the sum of the three segments.

[0172] Step 5: Optimize the existing path using a traversal optimization strategy to make the path length approach the optimal solution. Specifically:

[0173] Determine the traversal optimization parameters, L y L represents the current path length being searched. yLast The parameters are compared to the path length of the previous search. When L y <L yLast When the optimal path is L y When L y >L yLast At that time, the optimal path is still L. yLast Ultimately, the optimal path is achieved by repeatedly selecting the path until the maximum number of iterations is reached, at which point the optimal path is output. By using traversal optimization to continuously find a better path, the sampling range is narrowed and the sampling efficiency is improved by shortening the path length.

[0174] Simultaneously, the searched optimal path data is transmitted to the mobile robotic arm, which moves according to the searched optimal path. The overhead crane follows the mobile robotic arm, thus reducing the time cost required for the mobile robotic arm's path planning and achieving coordinated control between the overhead crane and the mobile robotic arm.

[0175] Step Six: Introduce a dynamic environment monitoring and real-time adjustment mechanism, specifically as follows:

[0176] (1) During the process of the mobile robotic arm moving along the optimal path, environmental changes are monitored in real time. Various types of sensors are deployed in the factory environment, and LiDAR and camera sensors are used for real-time monitoring to realize real-time monitoring of obstacles and changes in the environment. These sensors continuously collect environmental data, including information such as the position, shape and movement speed of obstacles, and transmit the data to the central processor through a wireless network. The central processor processes this data in real time and updates the environmental model to ensure safety and efficiency.

[0177] (2) When a new obstacle is detected or an existing obstacle is removed, the current path execution is immediately paused. The central processing unit analyzes the sensor data in real time to determine whether a new obstacle has appeared or an existing obstacle has been removed. Once an environmental change is detected, a decision is made immediately to pause the current path execution and send a pause command to the mobile robotic arm through a wireless communication mechanism to ensure the safety and accuracy of the operation.

[0178] (3) Rerun the path planning algorithm to generate a new obstacle avoidance path and update the optimal path data. After receiving the pause command, the central processing unit restarts the IBTO-RRT path planning algorithm, takes into account the latest environmental information, including newly appearing or removed obstacles, generates a new obstacle avoidance path based on the updated environmental data, optimizes it to ensure that the path is the optimal solution, and stores the new optimal path data for later use.

[0179] (4) The new optimal path data is transmitted to the mobile robotic arm to ensure that it can still complete the task efficiently in the dynamic environment. The new optimal path data is transmitted to the mobile robotic arm through wireless communication. After receiving the data, the robotic arm begins to move according to the new path. At the same time, the coordinated control between the crane and the mobile robotic arm is ensured so that the two can move synchronously. A feedback mechanism is established to monitor the execution of the mobile robotic arm in real time to ensure that it moves smoothly on the new path.

[0180] By deploying sensors such as LiDAR and cameras in the factory environment, dynamic environmental monitoring and real-time data transmission are achieved. The central processing unit can intelligently analyze sensor data and quickly respond to environmental changes, pausing the current path and replanning the optimal obstacle avoidance path. The new path data is transmitted to the mobile robotic arm via wireless communication to ensure its efficient execution of tasks. At the same time, it coordinates with the overhead crane for control and establishes a feedback mechanism to monitor the execution status, thereby achieving safe and flexible operation.

[0181] Based on this embodiment, simulation experiments were conducted on various path planning algorithms. The simulation environment is as follows: Figures 9 to 13 As shown, black solid circles of different sizes represent obstacles, with the starting point at (0, 0) and the ending point at (15, 12). The simulation data is shown in Table 1. A further simulation was then conducted in a complex path planning environment, as shown in the table below. Figures 14 to 18 As shown, Figure 14 This demonstrates the non-optimal path of the RRT algorithm. Figure 15 and Figure 16 The experimental results are shown for the RRT* and Informed-RRT* algorithms, respectively. Both have the advantage of asymptotic optimization. The Informed-RRT* algorithm further optimizes the path by establishing hyperelliptic sampling constraints. Figure 17 This demonstrates that the BITO-RRT* algorithm also exhibits asymptotic optimization characteristics; Figure 18 The IBTO-RRT* algorithm proposed in this embodiment first finds a feasible path through RRT, then establishes elliptical constraints, performs bidirectional growth and direct connection detection, and demonstrates the ability to quickly establish hyperelliptical sampling constraints and reduce invalid sampling points, thereby reducing time cost. In 30 simulation experiments, the five algorithms RRT, RRT*, Informed-RRT*, BITO-RRT* and IBTO-RRT* were compared in terms of path length, time cost and overall performance.

[0182] Qualitative results analysis: The RRT algorithm performs poorly in terms of path length, with a significantly higher average path length than other algorithms, and exhibits poor convergence. While RRT* and BITO-RRT* have some longer paths, their convergence is improved. Informed-RRT* shows poor convergence in complex environments, while the IBTO-RRT algorithm significantly outperforms the other three algorithms in terms of path length. Regarding time cost, the RRT algorithm has a significant advantage but is prone to outliers. The IBTO-RRT* algorithm has lower average and median time values ​​than Informed-RRT* and BITO-RRT*, and slightly lower than RRT. Overall, BITO-RRT* and IBTO-RRT algorithms have high success rates. IBTO-RRT* shows advantages in average path length, path length standard deviation, average time, and its standard deviation, especially in terms of length standard deviation.

[0183] like Figures 19 to 20 The qualitative results show that the path lengths of the RRT algorithm are relatively dispersed, with an average length significantly higher than other algorithms. While some path lengths of RRT* and BITO-RRT* are higher than the average, their convergence is improved. Informed-RRT* exhibits poor convergence in complex environments, while the IBTO-RRT algorithm significantly outperforms the other three algorithms in terms of path length. Figure 19 This section presents the results regarding time cost. The RRT algorithm has a significant advantage in time cost but is prone to outliers. The IBTO-RRT* algorithm has lower time average and median than Informed-RRT* and BITO-RRT*, and slightly lower than RRT*. Figure 20 In summary, the BITO-RRT* and IBTO-RRT* algorithms have high success rates. IBTO-RRT* shows certain advantages in terms of average path length, path length standard deviation, average time and its standard deviation. In particular, in terms of path length standard deviation, the path length standard deviations of the five algorithms are 4.084, 2.233, 0.615, 0.185 and 0.041, respectively, which can significantly reflect the advantages of IBTO-RRT*. Specific data are shown in Table 2.

[0184] Table 1 Simulation data under simple conditions

[0185]

[0186] Table 2 Simulation data under complex conditions

[0187]

[0188]

[0189] In summary, by utilizing the above-mentioned technical solution of the present invention, the present invention optimizes the path planning algorithm of the mobile robotic arm to achieve efficient path search, which can effectively improve the crane's ability to follow the mobile robotic arm, ensure the coordinated operation between the robotic arm and the crane during the transportation of heavy objects, and avoid the crane's movement speed exceeding the robotic arm's following ability due to insufficient path planning speed, thereby reducing transportation time and improving overall operation efficiency.

[0190] 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 for collaborative control of a crane-mounted mobile robotic arm, characterized in that, The path planning method includes: S1. Obtain the position information of the mobile robotic arm and the overhead crane based on the two-dimensional plan of the factory, combine the position information with the path planning algorithm to generate a feasible path, and calculate the path length of the feasible path; S2. Establish hyperelliptical intelligent constraint sampling conditions with the path length of feasible paths as the major axis, and generate elliptical constraints by obtaining sampling points based on the hyperelliptical intelligent constraint sampling conditions. S3. Combine the elliptical constraint with the bidirectional growth direct connection strategy to generate a direct connection path, and calculate the path length of the direct connection path. Obtain the optimal planning path based on the path length of the direct connection path and the optimization strategy. S31. Construct a forward growth tree and a reverse growth tree with the starting position and ending position of the mobile robotic arm as the root points respectively, and generate forward random sampling points and reverse random sampling points based on elliptical constraints using the forward growth tree and the reverse growth tree. S32. In the forward growth tree and the reverse growth tree, find the forward node and the reverse node whose distance from the forward random sampling point and the reverse random sampling point is a threshold, respectively. S33. Extend the forward node to the forward random sampling point by a preset length, and extend the reverse node to the reverse random sampling point by a preset length, and generate a new forward node and a new reverse node based on the extension results. S34. Generate a direct path based on the new forward node and the new reverse node. After determining the existence of obstacles, calculate the path length of the direct path and combine it with the optimization strategy to obtain the optimal planned path. S4. Generate driving instructions based on the optimal planning path to control the movement path of the mobile robotic arm and the overhead crane, and introduce a dynamic environment monitoring mechanism to monitor and adjust the movement status of the mobile robotic arm and the overhead crane in real time.

2. The path planning method for collaborative control of a crane moving robotic arm according to claim 1, characterized in that, The process of obtaining the position information of the mobile robotic arm and overhead crane based on the factory's two-dimensional plan view, combining the position information with a path planning algorithm to generate a feasible path, and calculating the path length of the feasible path includes: S11. Extract the two-dimensional plan of the factory to obtain the initial position and target position of the mobile robotic arm and the overhead crane, as well as the location information of obstacles inside the factory; S12. Set the initial position of the mobile robotic arm as the starting position and the target position as the ending position, and use the path planning algorithm to generate an initial search tree, defining the starting position as the root point of the search tree; S13. Generate random sampling points based on the starting and ending positions, find nodes in the search tree whose distance from the random sampling points is a threshold, and generate new nodes by extending them by a preset length in the direction from the nodes to the random sampling points. S14. Based on the obstacle location information, determine whether there is a collision between the new node and the node. If a collision occurs, discard the path. If no collision occurs, add the new node to the search tree and connect the random sampling point with the new node. S15. Determine the distance relationship between the new node and the endpoint, generate a feasible path based on the distance result, and record the path length of the feasible path.

3. The path planning method for collaborative control of a crane moving robotic arm according to claim 1, characterized in that, The step of establishing hyperelliptical intelligent constraint sampling conditions with the path length of feasible paths as the major axis, and generating elliptical constraints by obtaining sampling points based on hyperelliptical intelligent constraint sampling conditions includes: S21. Define a superellipse based on superellipse intelligent constraint sampling, and randomly sample points within the unit circle according to the sampling rules, then transfer the random points to the superellipse. S22. Determine the center point of the hyperellipse based on the starting and ending positions of the mobile robotic arm on the horizontal and vertical axes, and define the hyperellipse radius information using the path length of the feasible path and the straight-line distance from the starting and ending positions of the mobile robotic arm. Determine the position of the hyperellipse based on the radius information and the center. S23. Use the hyperellipse position to determine the hyperellipse constraint sampling parameters, and determine the sampling points inside the hyperellipse based on the hyperellipse constraint sampling parameters.

4. The path planning method for collaborative control of a crane moving robotic arm according to claim 1, characterized in that, The process of generating direct paths based on forward and reverse new nodes, calculating the path length of the direct paths after determining the existence of obstacles, and obtaining the optimal planned path by combining optimization strategies includes: S341. Connect the new forward node and the forward random sampling point to obtain the forward search path, and connect the new reverse node and the reverse random sampling point to obtain the reverse search path. Determine the existence of obstacles on the forward search path and the reverse search path respectively. S342. If there are obstacles, discard the corresponding path. If there are no obstacles, add the new forward node and the new reverse node to the forward growth tree and the reverse growth tree respectively, and connect the new forward node and the new reverse node to obtain a direct path. S343. Calculate the path lengths of the forward search path, the reverse search path, and the direct connection path respectively. At the same time, determine the traversal optimization parameters based on the optimization strategy, and obtain the optimal planned path based on the traversal optimization parameters and the path length.

5. The path planning method for collaborative control of a crane moving robotic arm according to claim 4, characterized in that, The formula for calculating the path length of the forward search path is: ； The formula for calculating the path length of the reverse search path is: ； The formula for calculating the path length of the directly connected path is: ； In the formula, L 1 represents the path length of the forward search path. L 2 indicates the path length of the reverse search path. L z This indicates the path length of the directly connected path. n Indicates the number of dimensions. d Represents Euclidean distance. P zi Indicates a positive node. P zi+1 Indicates the first i +1 positive node, P fi Indicates the reverse node. P fi+1 Indicates the first i +1 reverse node, P znew Indicates a new node in the positive direction. P fnew Indicates a new node in reverse order. i Indicates the number of nodes.

6. The path planning method for collaborative control of a crane moving robotic arm according to claim 1, characterized in that, The process of generating drive commands based on the optimal planned path to control the movement path of the mobile robotic arm and the overhead crane, and introducing a dynamic environment monitoring mechanism to monitor and adjust the movement status of the mobile robotic arm and the overhead crane in real time, includes: S41. Generate driving instructions based on the optimal planning path and send the driving instructions to the mobile robotic arm drive controller. The mobile robotic arm and the overhead crane move according to the optimal planning path. S42. Using sensors and radar within the factory, monitor the movement of the mobile robotic arm, as well as information about the factory environment and obstacles, and transmit the monitoring results to the central processing unit. S43. The central processing unit receives the monitoring results, identifies abnormal states, generates decision instructions based on the abnormal states to control the mobile robotic arm, and regenerates the planned path to control the movement state of the mobile robotic arm.

7. The path planning method for collaborative control of a crane moving robotic arm according to claim 6, characterized in that, The process of generating drive commands based on the optimal planned path and sending the drive commands to the mobile robotic arm drive controller, so that the mobile robotic arm and the overhead crane move according to the optimal planned path, includes: S411. Generate path points based on the optimal planning path, analyze the joint angles of each joint of the mobile robotic arm when executing the path points, and determine the target speed based on the joint angles and the initial joint angles of the mobile robotic arm. S412. Combine the joint angle with the target speed as the driving command for the mobile robotic arm, and send the driving command to the mobile robotic arm drive controller through the communication interface. S413 The mobile robotic arm drive controller receives drive commands to control the movement of the mobile robotic arm. During the movement, the overhead crane follows the mobile robotic arm.

8. The path planning method for collaborative control of a crane moving robotic arm according to claim 7, characterized in that, The process of generating path points based on the optimal planned path, analyzing the joint angles of each joint of the mobile robotic arm when executing path points, and determining the target speed based on the joint angles and the initial joint angles of the mobile robotic arm includes: S4111. Generate a grid map based on the factory's two-dimensional plan view, and generate path Voronoi points on the grid map according to the optimal planning path and Voronoi map generation rules. S4112. Traverse the path through the Voronoi points, select the intersection points, and connect adjacent intersections. Extract guide points based on the relationship between the lines seen at the intersection points and obstacles, and use the guide points as path points. S4113. Use computer simulation technology to establish a finite element model of the mobile robotic arm, and simulate the target joint angle of the mobile robotic arm when it moves along the path point in the working process according to inverse kinematics within the finite element model. S4114. Obtain the initial joint angles of the mobile robotic arm, and determine the joint angular velocity of the mobile robotic arm based on the initial joint angles and the target joint angles.

9. A path planning method for collaborative control of a crane moving robotic arm according to claim 8, characterized in that, The formula for calculating the joint angular velocity of the mobile robotic arm is as follows: ； In the formula, T This indicates the joint angular velocity of the moving robotic arm. k p Indicates proportional gain. T target Indicates the target joint angle. T cur This indicates the initial joint angle.

Images