Pneumatic soft mechanical arm obstacle avoidance trajectory planning method and system

Abstract

The invention provides a pneumatic soft mechanical arm obstacle avoidance trajectory planning method and system, and relates to the technical field of intelligent robot path planning and obstacle avoidance control, and the method comprises the steps that task space information of a target mechanical arm is acquired, and the task space information comprises starting point coordinates and target point coordinates of the tail end; the method comprises the steps of taking a starting point as a current node of a random tree, constructing an effective sampling space between the current node and a target point, generating a new sampling point in the effective sampling space by adopting a triple-guidance adaptive sampling strategy integrating target guidance, obstacle guidance and gradient guidance, constructing a new current node of the random tree by utilizing the new sampling point, and performing target guidance and obstacle guidance on the new current node. Iterative expansion of the random tree is realized until a target point is reached, and a preliminary collision-free path is obtained; the preliminary collision-free path is optimized, and a final collision-free path is obtained and used for controlling the obstacle avoidance track of the target mechanical arm from the starting point to the target point; according to the method, the path planning capability and the execution stability in a complex environment are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent robot path planning and obstacle avoidance control, and in particular to a pneumatic soft robotic arm obstacle avoidance trajectory planning method and system. Background Art

[0002] As a key piece of equipment in industrial production, robotic arms have garnered widespread attention from scholars across multiple disciplines. Currently, rigid robotic arms are the most widely used, deployed in logistics, industrial manufacturing, aerospace, and other scenarios. However, due to the inherent limitations of their structural form and material properties, traditional rigid robotic arms perform poorly in tasks requiring high flexibility, such as grasping fragile objects, assisting with surgical procedures, and assembling precision instruments. In recent years, with the technological breakthroughs and deep integration of disciplines such as bionics, materials science, and control science, researchers have developed a variety of soft robotic arms with biomimetic flexible structures, typically constructed from flexible materials such as silicone and resin. These arms offer significant advantages, including high degrees of freedom, strong human-machine interaction safety, and excellent adaptability to complex environments. Consequently, they hold great promise in a variety of applications, including clinical surgery, biological specimen collection, and precision instrument assembly. Compared to rigid robotic arms, soft robotic arms offer a wider range of actuation methods, including pneumatic, cable, and shape memory alloy actuation. Pneumatic actuation is currently one of the most widely used actuation methods, due to its readily available and environmentally friendly properties.

[0003] In recent years, the prototype design and trajectory planning of pneumatic soft robotic arms have attracted widespread attention from scholars in many fields. Although much progress has been made in the trajectory planning of pneumatic soft robotic arms, many challenges still exist in achieving efficient and collision-free motion in complex environments. Due to their flexibility and continuous deformation characteristics, pneumatic soft robotic arms are prone to unexpected contact or entanglement with obstacles during operation, so the obstacle avoidance problem is particularly critical and challenging. In response to this, scholars have proposed a variety of obstacle avoidance methods for different types of soft robotic arms. However, the collision detection of existing soft robotic arms is mostly based on the configuration space. However, this method is difficult to directly and in real time reflect the actual collision relationship between obstacles and the soft robotic arm joints, especially when the configuration space parameters do not change significantly. In addition, there is relatively little research on obstacle avoidance algorithms for pneumatic soft robotic arms.

[0004] Therefore, the existing obstacle avoidance trajectory planning technology for pneumatic soft robotic arms has problems such as poor path quality, low obstacle avoidance efficiency, and difficulty in accurately reflecting the actual collision relationship of joints. Summary of the Invention

[0005] In order to solve the above problems, the present invention proposes a pneumatic soft robotic arm obstacle avoidance trajectory planning method and system, which utilizes collision detection in task space, an adaptive sampling strategy guided by a triple strategy, an adaptive effective area expansion strategy, and a greedy optimization strategy to improve its path planning capability and execution stability in complex environments.

[0006] According to some embodiments, the present invention adopts the following technical solutions: A pneumatic soft robotic arm obstacle avoidance trajectory planning method, comprising: Acquire task space information of the target manipulator, wherein the task space information includes the starting point coordinates and the target point coordinates of the end; Taking the starting point as the current node of the random tree, an effective sampling space is constructed between the current node and the target point. A triple-guided adaptive sampling strategy that integrates target guidance, obstacle guidance, and gradient guidance is adopted to generate new sampling points in the effective sampling space. The new sampling points are used to construct a new current node of the random tree. The random tree is iteratively expanded until the target point is reached, and a preliminary collision-free path is obtained. The preliminary collision-free path is optimized to obtain the final collision-free path, which is used to control the obstacle avoidance trajectory of the target robotic arm from the starting point to the target point.

[0007] According to some embodiments, the present invention adopts the following technical solutions: A pneumatic soft robotic arm obstacle avoidance trajectory planning system, comprising: An acquisition module is configured to: acquire task space information of a target robotic arm, wherein the task space information includes a starting point coordinate and a target point coordinate of the end; The expansion module is configured to: use the starting point as the current node of the random tree, construct a valid sampling space between the current node and the target point, adopt a triple-guided adaptive sampling strategy that integrates target guidance, obstacle guidance, and gradient guidance, generate new sampling points in the valid sampling space, use the new sampling points to construct a new current node of the random tree, and iteratively expand the random tree until the target point is reached, obtaining a preliminary collision-free path; The optimization module is configured to optimize the preliminary collision-free path to obtain a final collision-free path for controlling the obstacle avoidance trajectory of the target robotic arm from the starting point to the target point.

[0008] According to some embodiments, the present invention adopts the following technical solutions: A computer program product includes a computer program, which implements the obstacle avoidance trajectory planning method of a pneumatic soft robotic arm when executed by a processor.

[0009] According to some embodiments, the present invention adopts the following technical solutions: A non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the method for planning the obstacle avoidance trajectory of a pneumatic soft robotic arm is implemented.

[0010] According to some embodiments, the present invention adopts the following technical solutions: An electronic device comprises: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the obstacle avoidance trajectory planning method of a pneumatic soft robotic arm.

[0011] Compared with the prior art, the present invention has the following beneficial effects: The trajectory planning method proposed in this paper improves upon the traditional configuration-space-based collision detection approach by adopting a collision detection mechanism in the task space. This method more accurately reflects the structural characteristics and dynamic deformation behavior of the pneumatic soft manipulator during actual operation. Furthermore, the multi-strategy fusion sampling mechanism and adaptive effective region update method make the path search process more efficient, significantly improving trajectory smoothness and planning quality. The overall method possesses strong generalization capabilities and algorithmic stability, effectively supporting high-precision obstacle avoidance planning tasks in complex scenarios.

[0012] The present invention can significantly improve the obstacle avoidance capability and trajectory execution efficiency of pneumatic soft robotic arms in unstructured environments, ensure global collision-free movement of the end and joints of the robotic arm during the execution of complex tasks, and improve operational safety; by introducing task space collision detection and a three-strategy sampling mechanism, the directionality and adaptability of path planning are enhanced, and the planning time is shortened; it demonstrates superior trajectory quality and execution effects in multiple types of simulations and experimental verifications, providing new ideas and feasible solutions for obstacle avoidance planning of flexible robots. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0014] Figure 1 Schematic diagram of the geometric structure of the four-cavity soft robotic arm of Example 1; Figure 2 This is a schematic diagram of the collision detection method proposed in Example 1; Figure 3 The figure shows the simulation results of the soft robotic arm posture and end path according to the method proposed in Example 1; Figure 4 The obstacle avoidance simulation of the method proposed in Example 1 is Plane projection result diagram; Figure 5 A structural diagram of the experimental platform built for Example 1; Figure 6 The experimental results of the soft robotic arm posture and end path of the method proposed in Example 1 are shown; Figure 7 The obstacle avoidance experiment of the method proposed in Example 1 is Plane projection result diagram; Figure 8 This is a schematic diagram of part of the obstacle avoidance process in the obstacle avoidance experiment of the method proposed in Example 1. DETAILED DESCRIPTION The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0015] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0016] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "comprising" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0017] Example 1 In one embodiment of the present invention, a method for planning an obstacle avoidance trajectory for a pneumatic soft robotic arm is provided, comprising: Step S1: Acquire the task space information of the target manipulator, wherein the task space information includes the starting point coordinates and the target point coordinates of the end; Step S2: With the starting point as the current node of the random tree, an effective sampling space is constructed between the current node and the target point. A triple-guided adaptive sampling strategy that integrates target guidance, obstacle guidance, and gradient guidance is used to generate new sampling points in the effective sampling space. The new sampling points are used to construct a new current node of the random tree. The random tree is iteratively expanded until the target point is reached and a preliminary collision-free path is obtained. Step S3: Optimize the preliminary collision-free path to obtain the final collision-free path, which is used to control the obstacle avoidance trajectory of the target manipulator from the starting point to the target point.

[0018] As an example, Figure 1Taking the four-cavity soft robotic arm with two curved segments, each containing four independent air cavities, as an example, a new obstacle avoidance trajectory planning method is proposed. This method uses a symmetrical motion mode to achieve motion decoupling of the curved segments, and designs a new collision detection method based on its kinematic model. For the first time, the physical structure and real-time deformation characteristics of the pneumatic soft robotic arm are taken into consideration. On this basis, in response to the shortcomings of the traditional RRT algorithm in complex environments, a three-strategy guided adaptive sampling strategy, an adaptive effective area expansion strategy, and a greedy optimization strategy are proposed to improve path planning efficiency and path quality. Combined with the above methods, a collision-free and high-quality trajectory path can be generated. Simulation and experimental results show that the obstacle avoidance trajectory planning method proposed in this embodiment can achieve good obstacle avoidance effect and has high practical application value. The specific implementation process is described in detail below, including the following parts: 1. Collision Detection Specific, definition and Surrounding the two curved segments The rotation angle of the axis, and Surrounding the two curved segments The rotation angle of the axis, and is the curvature radius of the two curved segments. In addition, the coordinate system of the initial plane of the first curved segment is defined as , the coordinate system of the end plane of the first curved segment is (also the initial plane of the second curved segment), the coordinate system of the end plane of the second curved segment is In addition, define The length of the central axis of the bending segment is , No. The first bending segment The length of the air cavity is , No. The first bending segment The radius of curvature of the air cavity is ,in , .

[0019] In order to decouple the motion of the two bending segments and control the position of the end gripper, a symmetrical motion pattern is imposed on the two bending segments, that is, the first bending segment The air cavity is relative to the second curved section Based on the imposed symmetric motion pattern, the following relationship can be obtained: (1) For obstacles of any shape and the two curved segments of the pneumatic soft robot arm, a sphere and a cylinder are used to envelop them, respectively. To perform collision detection, the minimum distance between the obstacle and the pneumatic soft robot arm must be calculated. Specifically, the distance between the center of the obstacle and each point on the central axis of the pneumatic soft robot arm's two curved segments is calculated to obtain the shortest distance. Collision detection is achieved by comparing this shortest distance with the radius of the obstacle and the radius of the pneumatic soft robot arm's cylindrical model.

[0020] Definition The radius of the obstacle is , No. The coordinates of the center of the obstacle are , the cylindrical radius of the two bending sections of the pneumatic soft robot arm is , the initial plane center position coordinates of the first curved segment are , the coordinates of the center position of the end of the first curved segment are , the coordinates of the center position of the end of the second curved segment are Analyze the motion characteristics of the four-cavity soft robotic arm and establish the relationship between the task space and configuration space of the two bending segments: (2) (3) (4) Combining equations (3) and (4), we can deduce (5) (6) Substituting formula (1) into formulas (2), (5), and (6), we can obtain (7) (8) (9) Will Substituting into equations (2), (5), and (6), we can obtain Corresponding configuration space parameters 、 、 .Will Substituting into equations (7)-(9), we can get Corresponding configuration space parameters 、 、 According to the geometric characteristics of the arc in three-dimensional space, any point on the arc is The rotation angle of the axis is equal to its radius of curvature. Therefore, based on equations (2), (5), and (6), the coordinates of any point on the central axis of the first curved segment can be obtained as: (10) (11) (12) in, Based on equations (1)-(9), the coordinates of any point on the central axis of the second curved segment can be obtained as: (13) (14) (15) in, .

[0021] Finally, by analyzing the motion characteristics of the four-cavity soft robotic arm and establishing the transformation relationship between the two curved segments from the task space to the configuration space, we can obtain the rotation angle, rotation angle, and curvature radius of the curved segment corresponding to the coordinates of the end center position:

[0022] The configuration space is the rotation angle, rotation angle and curvature radius of the curved segment, including the two curved segments around Axis rotation angle and , two curved segments around Axis rotation angle and , the curvature radius of the two curved segments and , the task space is the coordinates of the end center position, including the coordinates of the end center position of the first curved segment and the end center coordinates of the second curved segment .

[0023] Based on the transformation relationship between the two curved segments from the task space to the configuration space, we can further obtain the transformation relationship between the two curved segments from the configuration space to the task space, that is, the coordinates of any point on the curved segment:

[0024] in, , , represents any point on the first curved segment, Represents any point on the second curved segment.

[0025] Then, the distance between the obstacle center and each point on the two curved segments of the pneumatic soft robot arm is calculated. According to equations (10)-(15), we can obtain: (16) (17) in, represents the distance from any point on the first curved segment to the center of the obstacle sphere, Represents the distance from any point on the second curved segment to the center of the obstacle sphere.

[0026] Then, calculate the shortest distance from the obstacle center to the two curved segments, substitute equations (10)-(15) into equations (16) and (17), and calculate and , we can get: (18) (19) when hour, ;when hour, Therefore, substituting equation (18) into equation (16), we get The minimum value of Similarly, when hour, ;when hour, Therefore, substituting equation (19) into equation (17), we get The minimum value of .

[0027] In summary, considering the distance from the obstacle to any point on the two curved segments, calculate the shortest distance and use the collision distance threshold , and get the collision detection conditions: (20) in, Indicates the shortest distance from the obstacle to the first curved segment, represents the shortest distance from the obstacle to the second curved segment, Represents the sum of the obstacle radius and the curve radius.

[0028] 2. Collision-free path generation strategy Taking the starting point as the current node of the random tree, the effective sampling space between the current node and the target point is constructed. A triple-guided adaptive sampling strategy integrating target guidance, obstacle guidance, and gradient guidance is adopted to generate new sampling points in the effective sampling space. The new sampling points are used to construct a new current node of the random tree. The random tree is iteratively expanded until the target point is reached and a preliminary collision-free path is obtained. Specifically: (1) Triple-guided adaptive sampling strategy Given the limited workspace of the pneumatic soft robotic arm and to improve sampling efficiency and spatial exploration capabilities, this embodiment proposes a triple adaptive sampling strategy that integrates target guidance, obstacle guidance, and gradient guidance. The specific sampling method is as follows: (twenty one) in, is a random point, represents the probability threshold of target guided sampling, represents the probability threshold of obstacle-guided sampling, represents the sampling point, Represents the target point, represents the sampling points obtained by obstacle-guided sampling, represents the sampling points obtained by gradient-guided sampling.

[0029] To balance exploration and convergence during path planning, this embodiment adopts an adaptive strategy based on target distance to dynamically adjust the probability threshold of target-guided sampling, which is expressed as follows: (twenty two) in, represents the minimum value of the probability threshold for target guided sampling, represents the maximum value of the probability threshold for target guided sampling, Represents the current node of the random tree.

[0030] when , directly select the target point as the sampling point and guide the random tree to expand towards the target area. This strategy enhances the sampling density near the target, thereby accelerating the convergence of the algorithm, especially in the later stages of the search, and can significantly improve the path quality and planning efficiency.

[0031] when When , the obstacle-guided sampling strategy is triggered. In this strategy, a random obstacle center is selected as a reference point, and a directional offset with a random distance is applied around it to generate new sampling points. This mechanism can guide sampling near the obstacle boundary, effectively improving the coverage of spatial exploration and enhancing the algorithm's ability to traverse narrow channels. Accordingly, the expression for obstacle-guided sampling is: (twenty three) in, represents the randomly selected obstacle center point, is a randomly generated unit vector in three-dimensional space, is the offset distance sampled from a uniform distribution.

[0032] when When , the heuristic gradient guided sampling strategy is activated, which first randomly selects a reference point in the valid sampling area. , and then move a randomly selected step along the normalized direction towards the target point , to generate new sampling points. This strategy enhances the exploration ability towards the target while maintaining global sampling diversity, thereby improving the convergence speed and overall path quality. Correspondingly, the expression of the gradient-guided strategy is: (twenty four) in, It is the reference point uniformly sampled in the current valid area. is a randomly chosen step size.

[0033] In each iteration, the final sampling point is determined by a random number With threshold 、 Co-decision was made, resulting in a structured choice between the three sampling strategies.

[0034] (2) Adaptive effective area expansion strategy The collision-free trajectories generated by traditional RRT algorithms often lack clear directionality and may even extend in the opposite direction of the target, seriously affecting the efficiency of path planning. To address this issue, this embodiment introduces an adaptive sampling region update strategy into the traditional RRT algorithm to improve sampling guidance and overall planning performance.

[0035] First, define the starting point as , the target point is , the maximum number of iterations is , the step size is , the probability threshold of target guided sampling is , the probability threshold of obstacle-guided sampling is , new node and target point The minimum distance allowed between According to the starting point and the target point, the initial effective area can be obtained. axis, axis, The ranges of the axes are: (25) (26) (27) In the effective sampling area, the sampling points are obtained according to equations (21)-(24): , calculate each node on the random tree To the sampling point The Euclidean distance of , based on the following relationship, a arrive New node : (28) Where s is the step size.

[0036] Then, we can get and The parametric equation of the connection is: (29) in, .

[0037] According to formulas (2)-(10) and (29), detection and The collision between the paths is and The paths between them are collision-free. It will become a valid node. Then, the sampling area is updated by incorporating the new extended node into the valid sampling area as follows: (30) (31) (32) in, Represents a new node If and If there is a collision point between the paths, a new node is generated.

[0038] Then, the Euclidean distance between the newly generated valid node and the target point is calculated. When the following conditions are met, it is determined that the path has successfully reached the target point (33) If the Euclidean distance between the new node and the target point does not satisfy Equation (33), repeat the above steps until the target point is reached.

[0039] Based on the triple-guided adaptive sampling strategy and the adaptive effective area expansion strategy, a collision-free path is generated, and the nodes on the path are represented as follows: .

[0040] 3. Collision-free trajectory optimization strategy Although the adaptive sampling strategy and adaptive effective area expansion strategy based on triple guidance can initially generate a collision-free path, the path still has redundant nodes and bending inequality problems. In view of this, and to further improve the execution efficiency of the path, this embodiment designs a greedy optimization strategy for the initial collision-free path. To optimize, the specific steps are as follows: Step 3.1: Connect the starting point and target point , detect the collision of the connection.

[0041] If the line connecting the start point and the target point is collision-free, will be the final path.

[0042] If there is a collision point on the line connecting the starting point and the target point, discard the line connecting the starting point and the target point and connect and , and continue to detect collisions of the connections.

[0043] Repeat the above process in the reverse direction along the path until the first point without collision with the starting point is found, which is recorded as .

[0044] Step 3.2, As a new starting point, connect and , detect the collision of the connection.

[0045] like and The connection is collision-free. will be the final path.

[0046] like and There is a collision point on the connection line, discard this connection and connect and , and continue to detect collisions of the connections.

[0047] Repeat the above process until the first The points where the lines do not collide are recorded as .

[0048] Step 3.3: Follow the methods in steps 3.1 and 3.2 to continuously update the current starting point until it is successfully connected to the target point.

[0049] Finally, the path after greedy optimization is expressed as .

[0050] Compared to the initial collision-free path , While ensuring no collision, the number of path nodes is significantly reduced. 4. Trajectory tracking For the final optimal path , each node is connected by a straight line. Therefore, in order to ensure that the end of the pneumatic soft robot arm can accurately track the straight path, the linear interpolation method is used in the task space for trajectory planning. For example, based on the linear interpolation method, we can get (35) (36) (37) in, , Indicates from arrive Planning is time-consuming, represent The coordinates of represent The coordinates of express The coordinates of the moment.

[0051] Based on the transformation relationship between the two curved segments from task space to configuration space, we can further obtain the configuration space parameter trajectory of the pneumatic soft robotic arm from the starting point to the target point. By continuously adjusting the configuration space parameters, the pneumatic soft robotic arm eventually reaches the target point.

[0052] This embodiment simulates and verifies the effect of the obstacle avoidance trajectory planning method, specifically: 1. Simulation parameter settings To verify the obstacle avoidance capability of the proposed method in complex environments, a test was conducted in a simulation scenario containing multiple obstacles. The specific parameters were set as follows:

[0053] in, and represent the starting point and target point of the trajectory respectively, Indicates the center coordinates of the initial plane of the first curved segment. 、 、 、 、 represents the coordinates of the obstacle's center, 、 、 、 、 represents the radius of the obstacle, Represents the equivalent cylindrical radius of the two bending segments of the pneumatic soft robot arm.

[0054] 2. Experimental parameter settings To further verify the obstacle avoidance performance of the proposed method in complex environments, relevant tests were carried out in an experimental scenario containing multiple obstacles.

[0055] According to actual measurements, the radius of the five obstacles Both , the equivalent cylindrical radius of the two bending sections of the pneumatic soft robot arm for The rest of the parameters in the experimental test are set as follows:

[0056] 3. Description of simulation experiment results 3.1 Simulation Results To verify the feasibility of the obstacle avoidance trajectory planning proposed in this embodiment, a numerical simulation is first performed in the MATLAB / Simulink environment.

[0057] The simulation results are shown in the attached Figure 3 , Attachment Figure 4 As shown: Attachment Figure 3 In the figure, the black curve represents two curved segments, the colored balls represent obstacles, the blue solid line represents the collision-free path obtained based on the traditional RRT algorithm, and the red solid line represents the collision-free path obtained based on the obstacle avoidance trajectory planning method proposed in this embodiment.

[0058] Attachment Figure 4 In the figure, the black solid circle represents the obstacle in The blue solid line represents the collision-free path obtained based on the traditional RRT algorithm. The projection of the plane, the red solid line represents the collision-free path obtained based on the obstacle avoidance trajectory planning method proposed in this embodiment. Projection of a plane.

[0059] From the attached Figure 2 and 3It can be seen that the generated optimized path remains collision-free throughout, while significantly reducing path tortuosity. Compared to the traditional RRT algorithm, the trajectory planning method proposed in this embodiment effectively eliminates redundant nodes in the path, resulting in a smoother and more continuous trajectory with higher motion efficiency. In addition, the two curved segments of the pneumatic soft robotic arm successfully avoided all obstacles during the path planning process, further verifying the effectiveness of this embodiment in achieving high-precision obstacle avoidance in complex environments.

[0060] It is worth emphasizing that in an environment where obstacles are densely distributed and numerous, the algorithm proposed in this embodiment can still generate a collision-free path with a smaller number of nodes, fully demonstrating its high adaptability and robustness to narrow and cluttered spatial environments.

[0061] 3.2 Experimental Results In order to further verify the practical application effect of the proposed collision-free trajectory planning method, the attached Figure 5 The experimental platform shown in the figure. The platform mainly includes: a host computer running MATLAB / Simulink RTWT, a real-time target controller, a terminal board, eight pneumatic proportional valves, four sets of DC regulated power supplies, an air compressor, a four-cavity soft robotic arm, and multiple obstacles. The control algorithm runs in the host computer and transmits the control instructions to the real-time target controller via Ethernet; then, the real-time controller outputs the signal to the terminal board to drive each pneumatic proportional valve and adjust the air pressure in each air cavity to achieve precise control of the motion state of the soft robotic arm. At the same time, the feedback voltage signal of the pneumatic valve is returned to the host computer via the terminal board and the real-time controller to achieve closed-loop control and data acquisition. The air compressor continuously provides a stable air source for the pneumatic system, and the four sets of DC power supplies output a constant In the experiment, obstacles were suspended from a support frame by thin wires to simulate the complex obstacle layout in an actual working environment, thereby comprehensively testing the obstacle avoidance capability and stability of the proposed algorithm in dynamic real-world scenarios.

[0062] The results of the experiment are shown in the attached Figure 6 , Attachment Figure 7 , Attachment Figure 8 shown.

[0063] Attachment Figure 6 In the figure, the black curve represents the simulated posture of the two curved segments, the transparent figure represents the actual posture of the two curved segments, the green transparent ball represents the obstacle, the blue solid line represents the collision-free path obtained based on the traditional RRT algorithm, and the red solid line represents the collision-free path obtained based on the obstacle avoidance trajectory planning method proposed in this embodiment.

[0064] Attachment Figure 7In the figure, the green solid circle represents the obstacle in The blue solid line represents the collision-free path obtained based on the traditional RRT algorithm. The projection of the plane, the red solid line represents the collision-free path obtained based on the obstacle avoidance trajectory planning method proposed in this embodiment. The blue trajectory contains nodes, and the path length is ; while the red track contains only nodes, and the path length is Compared with the traditional RRT algorithm, the improved algorithm proposed in this embodiment reduces the number of nodes while ensuring that the path is collision-free. , the path length is shortened , which significantly reduces the tortuosity and redundancy of the trajectory, fully demonstrating the advantages of the method proposed in this embodiment in terms of path simplicity and planning efficiency.

[0065] Attachment Figure 8 In the diagram, the arrows indicate the order of the diagrams, and the points and As shown in Figure 8, both bending segments of the pneumatic soft robot arm successfully avoided obstacles during the movement, fully verifying the effectiveness and reliability of the proposed collision detection method in complex environments.

[0066] In summary, the obstacle avoidance trajectory planning method proposed in this embodiment can realize efficient obstacle avoidance and precise path generation of the pneumatic soft robotic arm in complex environments. It not only effectively improves the path smoothness and planning efficiency, but also significantly enhances the system's adaptability to narrow spaces and dense obstacles, demonstrating excellent planning performance and engineering application potential. Example 2 In one embodiment of the present invention, a pneumatic soft robotic arm obstacle avoidance trajectory planning system is provided, comprising: An acquisition module is configured to: acquire task space information of a target robotic arm, wherein the task space information includes a starting point coordinate and a target point coordinate of the end; The expansion module is configured to: use the starting point as the current node of the random tree, construct a valid sampling space between the current node and the target point, adopt a triple-guided adaptive sampling strategy that integrates target guidance, obstacle guidance, and gradient guidance, generate new sampling points in the valid sampling space, use the new sampling points to construct a new current node of the random tree, and iteratively expand the random tree until the target point is reached, obtaining a preliminary collision-free path; The optimization module is configured to optimize the preliminary collision-free path to obtain a final collision-free path for controlling the obstacle avoidance trajectory of the target robotic arm from the starting point to the target point.

[0067] Example 3 In one embodiment of the present invention, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the aforementioned method for obstacle avoidance trajectory planning of a pneumatic soft robotic arm.

[0068] Example 4 In one embodiment of the present invention, a non-transitory computer-readable storage medium is provided, which is used to store computer instructions. When the computer instructions are executed by a processor, the pneumatic soft robotic arm obstacle avoidance trajectory planning method is implemented.

[0069] Example 5 In one embodiment of the present invention, an electronic device is provided, comprising: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes the pneumatic soft robotic arm obstacle avoidance trajectory planning method.

[0070] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0071] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0072] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. A pneumatic soft robot arm obstacle avoidance trajectory planning method, characterized in that: include: Acquire task space information of the target manipulator, wherein the task space information includes the starting point coordinates and the target point coordinates of the end; Taking the starting point as the current node of the random tree, an effective sampling space is constructed between the current node and the target point. A triple-guided adaptive sampling strategy that integrates target guidance, obstacle guidance, and gradient guidance is adopted to generate new sampling points in the effective sampling space. The new sampling points are used to construct a new current node of the random tree. The random tree is iteratively expanded until the target point is reached, and a preliminary collision-free path is obtained. The preliminary collision-free path is optimized to obtain the final collision-free path, which is used to control the obstacle avoidance trajectory of the target robotic arm from the starting point to the target point.

2. The method for planning an obstacle avoidance trajectory of a pneumatic soft robot arm according to claim 1, wherein: The effective sampling space between the current sampling point and the target point is constructed by adopting an adaptive effective area expansion strategy, which is expressed as follows: in, express The effective area of ​​the axis, express The effective area of ​​the axis, express The effective area of ​​the axis, and Represent the coordinates of the starting point and the target point respectively, Indicates the coordinates of the newly generated valid node.

3. The method for planning an obstacle avoidance trajectory of a pneumatic soft robot arm according to claim 1, wherein: The triple-guided adaptive sampling strategy integrating target guidance, obstacle guidance, and gradient guidance is expressed as follows: in, represents the sampling point, is a random point in the valid sampling space, represents the probability threshold of target guided sampling, represents the probability threshold of obstacle-guided sampling, Represents the target point, represents the sampling points obtained by obstacle-guided sampling, represents the sampling points obtained by gradient-guided sampling.

4. The method for planning an obstacle avoidance trajectory of a pneumatic soft robot arm according to claim 3, wherein: The probability threshold of the target guided sampling adopts an adaptive strategy based on the target distance to dynamically adjust the probability threshold of the target guided sampling, and its expression is as follows: in, represents the minimum value of the probability threshold for target guided sampling, represents the maximum value of the probability threshold for target guided sampling, Indicates the current node.

5. The method for planning an obstacle avoidance trajectory of a pneumatic soft robot arm according to claim 1, wherein: The new sampling point is used to construct a new current node of the random tree, specifically: Based on the new sampling point , select a node from the random tree with the shortest Euclidean distance to the sampling point ; Generate a arrive New node ; Through the collision detection of the task space, the end of the robot arm is detected arrive The collision situation with obstacles in the path, if there is no collision, the node Become a new valid node in the random tree, and regenerate a new node if there is a collision .

6. The method for planning an obstacle avoidance trajectory of a pneumatic soft robot arm according to claim 5, wherein: The collision detection of the task space is specifically as follows: Based on the transformation relationship of the bending segment from the task space to the configuration space, the target robot arm end is calculated from arrive The rotation angle, rotation angle and curvature radius of the curved segment in the path; Based on the transformation relationship of the bending segment from the configuration space to the task space, the target robot arm end is calculated from arrive The coordinates of any point on the curved segment of the path; Calculate the distance from the obstacle to any point on the curved segment; Collision detection is performed based on a pre-calculated collision distance threshold.

7. The method for planning an obstacle avoidance trajectory of a pneumatic soft robot arm according to claim 6, wherein: The optimization of the preliminary collision-free path is to perform task space collision detection on the paths between nodes in the collision-free path through a greedy optimization strategy, thereby reducing redundant nodes on the collision-free path and reducing the tortuosity of the path.

8. A pneumatic soft robot arm obstacle avoidance trajectory planning system, characterized in that: include: An acquisition module is configured to: acquire task space information of a target robotic arm, wherein the task space information includes a starting point coordinate and a target point coordinate of the end; The expansion module is configured to: use the starting point as the current node of the random tree, construct a valid sampling space between the current node and the target point, adopt a triple-guided adaptive sampling strategy that integrates target guidance, obstacle guidance, and gradient guidance, generate new sampling points in the valid sampling space, use the new sampling points to construct a new current node of the random tree, and iteratively expand the random tree until the target point is reached, obtaining a preliminary collision-free path; The optimization module is configured to optimize the preliminary collision-free path to obtain a final collision-free path for controlling the obstacle avoidance trajectory of the target robotic arm from the starting point to the target point.

9. A non-transitory computer-readable storage medium, characterized in that The non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by the processor, a pneumatic soft robotic arm obstacle avoidance trajectory planning method as described in any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: include: A processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement a pneumatic soft robotic arm obstacle avoidance trajectory planning method as described in any one of claims 1 to 7.

Images