Underwater double-arm robot target grabbing coordinated planning method and system based on double-population cascade hierarchical particle swarm optimization

Abstract

The invention discloses an underwater double-arm robot target grabbing coordinated planning method and system based on double-population cascade hierarchical particle swarm optimization, and belongs to the field of underwater operation automatic control. A multi-target motion planning model of the underwater double-arm robot is constructed based on two levels of a local optimization task of symmetrical cooperation of double mechanical arms and a global optimization task of division cooperation of a robot hull and the mechanical arms; a double-arm-double-group double-population structure is constructed; the exploration population A and the development population B perform information interaction between an upper layer and a lower layer through an information sharing set; checking whether iteration meets a convergence condition or not; if the maximum number of iterations is reached or the change of the fitness value is smaller than a preset threshold value, stopping iteration and outputting a globally optimal solution; otherwise, returning to continue iteration of the double-population value. The method is used for solving the problems that the high-dimensional complex optimization of the underwater double-arm robot in the underwater complex environment falls into local optimization, the multi-target weighing is poor, and the efficiency between global search and local optimization is unbalanced.

Description

Technical Field

[0001] This invention belongs to the field of underwater operation automatic control, specifically relating to a target grasping coordination planning method and system for an underwater dual-arm robot based on dual-population hierarchical particle swarm optimization. Background Technology

[0002] In recent years, with the increasing complexity and diversity of underwater operations, the demand for robots with greater flexibility, stability, and high-precision collaborative operation capabilities has become increasingly urgent. Underwater dual-arm robots, due to their superior collaborative operation capabilities and adaptability to complex tasks, have gradually become one of the hot topics in underwater robotics research. Compared to traditional single-arm underwater robot systems, dual-arm systems demonstrate advantages in performing complex underwater tasks. For example, dual arms can collaboratively grasp large objects, perform symmetrical assembly, and carry out precision operations, playing an irreplaceable role in underwater repair, sample collection, and equipment installation scenarios. Furthermore, dual-arm systems also exhibit unique advantages in task reliability and redundancy: when one arm is limited or malfunctions, the other arm can take over part of the task, thereby reducing the risk of task failure.

[0003] Despite the significant advantages of underwater dual-arm robots in terms of task execution efficiency and flexibility, their motion planning problem still faces numerous challenges, especially in the complex underwater environment where balancing system stability, path efficiency, and disturbance suppression is crucial. Existing motion planning algorithms still have significant shortcomings. For example, traditional single-swarm particle swarm optimization algorithms are prone to getting trapped in local optima in high-dimensional complex optimization problems and have limited ability to handle multi-objective trade-offs. Furthermore, the lack of effective hierarchical search strategies leads to an efficiency imbalance between global search and local optimization, thereby reducing the quality of solutions and the adaptability of the algorithm.

[0004] Therefore, there is an urgent need for a coordinated planning method applicable to high-dimensional complex optimization problems of underwater dual-arm robots. Summary of the Invention

[0005] This invention provides a target grasping coordination planning method and system for underwater dual-arm robots based on dual-swarm hierarchical particle swarm optimization, which is used to solve the problems of underwater dual-arm robots getting trapped in local optima in high-dimensional complex optimization in complex underwater environments, poor trade-offs for multiple objectives, and efficiency imbalance between global search and local optimization.

[0006] This invention is achieved through the following technical solution: A coordinated planning method for target grasping of an underwater dual-arm robot based on dual-population hierarchical particle swarm optimization, the method comprising the following steps: Step S1: Construct a multi-objective motion planning model for the underwater dual-arm robot based on two levels: local optimization task of symmetrical cooperation between the two robotic arms and global optimization task of division of labor and cooperation between the robot hull and the robotic arms. Clarify the corresponding mechanical safety constraints and system safety constraints. Step S2: Construct a dual-population structure of "two arms - two groups", including exploratory population A and development population B; Step S3: Based on the dual-population structure established in step S2, the corresponding hierarchical search framework is as follows: The L1 layer, as a wide-area search layer, covers the entire solution space to cope with the complexity of high-dimensional multi-objective optimization in two arms. Population A is responsible for discovering candidate solutions that satisfy global safety and stability. The L2 layer, as a local fine-grained search layer, limits the scope to the high-quality region passed by the L1 layer, and the population B discovers candidate solutions that satisfy the multi-task constraints and local accuracy requirements of the two arms. Step S4: The exploration population A and the development population B interact with each other through an information sharing set between the upper and lower layers. During each iteration, the two populations write their respective Pareto high-quality solutions into the sharing set and obtain optimization information from each other to update their own search directions. Step S5: Check whether the iteration in step S4 meets the convergence condition; if the maximum number of iterations is reached or the fitness value change is less than the preset threshold, terminate the iteration and output the global optimal solution; otherwise, return to continue iterating the bipopulation value.

[0007] Furthermore, the multi-objective motion planning model in step S1 includes a path length objective function, a joint change minimization objective function, and the disturbance of the hull by the manipulator's motion; The path length objective function reflects the total movement distance required for the underwater dual-arm robot to complete its task and the effectiveness of the functional division between the hull and the dual arms. It is defined as the sum of the Euclidean distances of the path point sequence:

[0008] In the formula, Indicates the first The three-dimensional coordinates of each path point; This represents the total number of path points. Represents the total path length of the entire motion process; Specifically, the objective function for minimizing joint changes is defined as the weighted sum of the changes in the joint angles of the robotic arms:

[0009] In the formula, Indicates the first The change in the angle of each joint; These are the weighting coefficients for the changes in the angles of each joint; Indicates the stability index of the system; The disturbance to the hull caused by the manipulator's motion is specifically as follows: Considering symmetry, the objective function of the restoring torque of the underwater dual-arm robot during its motion is:

[0010] In the formula, The total restoring torque generated by the left arm; The total restoring torque generated by the right arm; For the robotic arm in The restoring torque component in the direction, for the left arm is The right arm is .

[0011] Furthermore, the safety constraints in step S1 include mechanical safety constraints and system safety constraints, specifically as follows: Safety constraints on bi-arm joint angles, velocities, and accelerations, as well as integrated visual field-obstacle avoidance constraints; Risk constraints related to the bottoming out of the arms and the entire hull:

[0012] In the formula, This indicates the real-time vertical distance between the underwater dual-arm robot and the seabed. , The upper and lower threshold values ​​for this distance are defined respectively.

[0013] Furthermore, the obstacle avoidance safety constraints of the underwater dual-arm robot system and the mutual obstacle avoidance constraints between the two arms are specifically as follows: Using the polygon method, obstacles are enclosed by spheres in three-dimensional space, while the links and hull of the underwater dual-arm robot are simplified to cylinders. The three-dimensional model is converted into a two-dimensional model through mathematical mapping. The underwater dual-arm robot meets the minimum safety distance requirements. :

[0014] In the formula, The shortest distance from the obstacle to each segment of the robotic arm. This represents the actual minimum distance between any part of an underwater dual-arm robot and an obstacle. For a predefined safe distance threshold; Define the minimum safe distance constraint between the two robotic arms as follows:

[0015] In the formula, This represents the actual minimum distance between the two robotic arms. The preset safety threshold; When calculating the minimum distance between the two arms, consider the distances between all link pairs:

[0016] In the formula, For the left arm The first link and the right arm The minimum distance between links.

[0017] Furthermore, in step S2, the exploration population A specifically undertakes the global discovery task of the Pareto front, focusing on the global optimization of the overall coordination between the hull and the two arms, and is responsible for finding potential non-dominated solution regions in the entire target space, as shown in the equation:

[0018] in:

[0019] In the formula, Represents population size, For the first The formula utilizes the position of each particle to fully leverage information from all particles in the population, including information used in the traditional PSO algorithm. Item and Term, while retaining the acceleration coefficient and To maintain the basic characteristics of the algorithm; when When, it indicates that the particle Relative to particles It has a better fitness value; For any two chosen attractors or repulsions and , Its information interaction expression is:

[0020] In the formula, the adaptive random coefficients , Randomly select values ​​from the interval [-0.05, 0.05]. Indicates the current iteration number. Maximum number of iterations For control parameters; Weighting coefficient The calculation is performed in a normalized form:

[0021] In the formula, and They represent particles respectively and Fitness value, weight coefficient Then take as This forms an asymmetric weight pair. It is a very small positive number; The proposed velocity update rule leads to an improved particle evolution formula: .

[0022] Furthermore, the development population B in step S2 specifically focuses on fine-tuning the Pareto front region discovered by the exploration population, particularly satisfying end-grasping accuracy, arm spacing, and local constraints. By introducing an improved differential mutation strategy and an adaptive search mechanism, the accuracy and quality of the Pareto solution are effectively improved. Based on the search progress ratio... The following adaptive search direction adjustment mechanism was designed:

[0023] Search direction in the formula Through historical direction and the current optimal direction The weighted combination was obtained, realizing the cumulative utilization of search experience.

[0024] Based on the above mechanism, the particle position update rule is as follows:

[0025] In the formula, Given the current search radius, exponential decay ensures that the search accuracy gradually improves with each iteration; Adaptive learning rate; For the search direction, It is a function that generates standard normally distributed random numbers; The complete population B update strategy is as follows: .

[0026] Furthermore, the dual-population information interaction mechanism in step S4 specifically involves the following: during each iteration, both populations write their respective Pareto high-quality solutions into the information sharing set, and simultaneously obtain search information provided by the other to optimize their own evolutionary direction; an adaptive probability control mechanism based on fitness differences is employed.

[0027] In the formula, It is the probability of information exchange between populations. This represents the difference in fitness between the current optimal solutions of two populations. and These are adaptive control parameters; and The parameter values ​​will be dynamically adjusted during the iteration process.

[0028] An underwater dual-arm robot target grasping coordination planning system based on dual-population hierarchical particle swarm optimization is provided. The system uses the underwater dual-arm robot target grasping coordination planning method based on dual-population hierarchical particle swarm optimization as described above. The system includes a controller and its built-in planning and optimization module. The planning and optimization module is used to establish a multi-objective motion planning model for the underwater dual-arm robot and run a dual-population hierarchical particle swarm optimization algorithm to generate coordinated planning results. Based on the coordination planning results, the controller drives the dual-arm robot to autonomously complete target grasping and collaborative operation tasks.

[0029] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the method described above.

[0030] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described above.

[0031] The beneficial effects of this invention are: In complex underwater environments, the balance between stability, path efficiency, and disturbance suppression of underwater dual-arm robots is particularly critical. This invention mainly improves the coordinated movement of the hull and the two arms, achieving a good balance between three objectives: path length, joint variation, and hull disturbance. This invention provides a reliable motion planning scheme for the application of underwater dual-arm robots in complex underwater operation scenarios. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0033] Figure 2 This is a schematic diagram of the application process in an embodiment.

[0034] Figure 3 This invention describes the autonomous operation and movement process of an underwater dual-arm robot in an obstacle environment.

[0035] Figure 4 This is a mathematical model representing the collision relationship between the obstacle and the robot hull and robotic arm of the present invention, where (a) represents the disjoint relationship, (b) represents the tangent relationship, and (c) represents the intersecting relationship.

[0036] Figure 5 This is a schematic diagram illustrating the approach of the field of view to the target in this invention.

[0037] Figure 6 This is a schematic diagram of the hierarchical search framework of the present invention.

[0038] Figure 7 The present invention provides the objective function convergence process under obstacle conditions, wherein (a) is the path length objective function convergence process, (b) is the system disturbance objective function convergence process, and (c) is the joint change convergence process.

[0039] Figure 8 The motion planning trajectory of the underwater dual-arm robot in an obstacle environment according to the present invention. Detailed Implementation

[0040] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of this application with unnecessary detail.

[0041] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0042] It should also be understood that the terminology used in this application specification is for the purpose of describing particular embodiments only and is not intended to limit the application. As used in this application specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0043] The following is in conjunction with the appendix to this application specification. Figure 1-8 The technical solutions in the embodiments of this application are clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0044] Many specific details are set forth in the following description in order to provide a full understanding of this application. However, this application may also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the spirit of this application. Therefore, this application is not limited to the specific embodiments disclosed below.

[0045] Implementation Method 1 This embodiment provides a coordinated planning method for target grasping of an underwater dual-arm robot based on dual-population hierarchical particle swarm optimization, such as... Figure 1 As shown, the method includes the following steps: Step S1: Construct a multi-objective motion planning model for the underwater dual-arm robot based on two levels: local optimization task of symmetrical cooperation between the two robotic arms and global optimization task of division of labor and cooperation between the robot hull and the robotic arms. Clarify the corresponding mechanical safety constraints and system safety constraints. Step S2: Construct a dual-population structure of "two arms - two groups", including exploratory population A and development population B; Step S3: Based on the dual-population structure established in step S2, the corresponding hierarchical search framework is as follows: The L1 layer, as a wide-area search layer, covers the entire solution space to cope with the complexity of high-dimensional multi-objective optimization in two arms. Population A is responsible for discovering candidate solutions that satisfy global safety and stability. The L2 layer, as a local fine-grained search layer, limits the scope to the high-quality region passed by the L1 layer, and the population B discovers candidate solutions that satisfy the multi-task constraints and local accuracy requirements of the two arms. Step S4: The exploration population A and the development population B interact with each other through an information sharing set between the upper and lower layers. During each iteration, the two populations write their respective Pareto high-quality solutions into the sharing set and obtain optimization information from each other to update their own search directions. Step S5: Check whether the iteration in step S4 meets the convergence condition; if the maximum number of iterations is reached or the fitness value change is less than the preset threshold, terminate the iteration and output the global optimal solution; otherwise, return to continue iterating the bipopulation value.

[0046] This embodiment describes an underwater dual-arm robot system performing plug-in / plug-out docking operations in an environment with obstacles, such as... Figure 2 As shown, the underwater dual-arm robot platform consists of an underwater robot hull and two four-degree-of-freedom manipulators. To meet the requirements of underwater operations, the underwater robot is equipped with a propulsion system consisting of two main thrusters, two side thrusters, and four vertical thrusters. It also integrates navigation sensors such as a Doppler velocity gauge (DVL), altimeter, magnetic compass, and acoustic source localization sensor to acquire its position and attitude information in real time. In docking and insertion operations with obstacles, the dual-arm underwater dual-arm robot system is required to complete precise docking and insertion tasks while avoiding obstacles.

[0047] The application scheme of this embodiment is as follows: Figure 3 As shown, the main processes include: First, the population is randomly initialized, and the multi-objective function value of each particle is calculated. Second, in layer L1, the development and exploration population A performs a global Pareto front exploration task across the entire objective space. Simultaneously, in layer L2, the development population B extracts information about Pareto optimal solution regions from the information sharing set and performs local fine-tuning optimization within these regions. Third, the populations are updated and information interaction occurs: the exploration population A and the development population B interact through the information sharing set, updating their search directions and particle positions. Finally, convergence is checked: the algorithm is evaluated to see if it meets the convergence criteria. If the maximum number of iterations is reached or the fitness value change is less than a preset threshold, the algorithm terminates and outputs the global optimal solution; otherwise, it returns to continue iterating over the two population values.

[0048] Specifically, the multi-objective function value for each particle is calculated, including: Calculating the multi-objective function values ​​includes minimizing path length, joint variation, and the disturbance of the hull caused by the manipulator's motion: The objective function for the path length of an underwater dual-arm robot is defined as the sum of the Euclidean distances of the path point sequence:

[0049] In the formula, Indicates the first The three-dimensional coordinates of each path point; This represents the total number of path points. It represents the total path length of the entire motion process.

[0050] The objective function for minimizing the joint changes of an underwater dual-arm robot is defined as a weighted sum of the changes in the joint angles of the robotic arms:

[0051] In the formula, Indicates the first The change in the angle of each joint; These are the weighting coefficients for the changes in the angles of each joint; This indicates the stability index of the system.

[0052] For underwater dual-arm robot systems, considering the left and right robotic arms ( Indicates the left arm. Given the symmetrical arrangement of the right arm (representing the center of the hull), the total restoring torque generated by the system can be expressed as the vector superposition of the restoring torques generated by the two robotic arms:

[0053] In the formula, The total restoring torque generated by the left arm; This represents the total restoring torque generated by the right arm.

[0054] For the One robotic arm ( ), No. The restoring force generated by each link can be expressed as:

[0055] In the formula, and The first The first robotic arm The mass and buoyancy of each link, Indicates the first The rotation matrix from the link coordinate system to the hull coordinate system.

[0056] No. The expression for the restoring torque generated during the movement of each link in the hull coordinate system is:

[0057] In the formula, and The first The first robotic arm The positions of the center of gravity and center of buoyancy of each link in the hull coordinate system. For the first The transformation matrix from the link coordinate system to the hull coordinate system and The first The first robotic arm The position of the center of gravity and center of buoyancy of each link in its own frame of reference.

[0058] Considering symmetry, the objective function of the restoring torque of the underwater dual-arm robot during its motion is:

[0059] In the formula, The total restoring torque generated by the left arm; The total restoring torque generated by the right arm; For the robotic arm in The restoring torque component in the direction.

[0060] During population initialization, mechanical safety constraints and system safety constraints are clearly defined. Specifically, these include mechanical safety constraints for the dual-arm system, bottom-touching risk constraints, obstacle avoidance safety constraints for the underwater dual-arm robot system, mutual obstacle avoidance constraints between the two arms, and binocular vision constraints. The mechanical safety constraints of the dual-arm system are as follows:

[0061] In the formula, , , For each robotic arm joint at Angle, velocity, and acceleration at any given moment; , , These represent the angle, angular velocity, and extreme values ​​of angular acceleration for each joint.

[0062] Furthermore, the specific constraints regarding the risk of hitting bottom are as follows:

[0063] In the formula, This indicates the real-time vertical distance between the underwater dual-arm robot and the seabed. , The upper and lower threshold values ​​for this distance are defined respectively.

[0064] The specific obstacle avoidance safety constraints for the underwater dual-arm robot system are as follows: like Figure 4 As shown, there are three spatial relationships between the system and obstacles: disjoint, tangent, and intersecting. To ensure safe operation, the underwater dual-arm robot must remain disjoint or tangent to all obstacles, i.e., meet the minimum safe distance requirement:

[0065] In the formula, The shortest distance from the obstacle to each segment of the robotic arm. This represents the actual minimum distance between any part of an underwater dual-arm robot and an obstacle. This is a predefined safe distance threshold.

[0066] The mutual obstacle avoidance constraint between the two arms is specifically as follows: Define the minimum safe distance constraint between the two robotic arms:

[0067] In the formula, This represents the actual minimum distance between the two robotic arms. This is a preset safety threshold.

[0068] When calculating the minimum distance between the two arms, consider the distances between all link pairs:

[0069] In the formula, For the left arm The first link and the right arm The minimum distance between links.

[0070] like Figure 5As shown, the combined field-of-view and obstacle avoidance constraint is as follows: During the underwater dual-arm robot's approach to the target object via the hull movement, it continuously uses the target object's 3D coordinates fed back by the binocular cameras as navigation signals for its movement. Therefore, the target object must remain within the binoculars' field of view throughout the approach process.

[0071] In the formula, and These are the hull coordinates in the hull coordinate system. direction and The range of motion in the direction, and The target object in the hull coordinate system shaft and The coordinates of the axis, and To take into account the current hull height and obstacle distribution The maximum coordinates of the image plane in the hull coordinate system. and To take into account the current hull height and obstacle distribution The minimum coordinates of the image plane in the hull coordinate system.

[0072] Secondly, in the L1 layer, population A is explored, and a Pareto front global exploration task is performed on the entire target space. Further, the exploration of population A (global search) specifically involves: The exploratory population undertakes the global discovery task of the Pareto front, responsible for finding potential non-dominated solution regions throughout the entire target space. Pareto-optimal solutions guide the population towards these non-dominated regions, while dominated solutions identify target space regions that need to be avoided. Based on this idea, the following velocity update rule is proposed to fully utilize the information carried by all particles in the swarm, as shown in the following equation.

[0073]

[0074] in:

[0075] In the formula, Represents population size, For the first The position of each particle enables full utilization of information from all particles in the population, including information used in the traditional PSO algorithm. Item and Term, while retaining the acceleration coefficient and To maintain the basic characteristics of the algorithm. When When, it indicates that the particle Relative to particles It has a better fitness value. For any two chosen attractors (repellers)... and Its information interaction expression is:

[0076] Adaptive random coefficients , Randomly select values ​​from the interval [-0.05, 0.05]. Indicates the current iteration number. Maximum number of iterations This is a control parameter (empirical value taken as 0.8). Weighting coefficient. The calculations were performed in a standardized form: ,in and They represent particles respectively and Fitness value, weight coefficient Then take as This constitutes an asymmetric weight pair. It is a very small positive number (taken as 0.0001).

[0077] The proposed velocity update rule leads to an improved particle evolution formula, as shown in the equation:

[0078] Furthermore, the development of population B (local optimization) specifically involves: Population B focuses on fine-tuning the Pareto front region discovered by the exploration population. By introducing an improved differential mutation strategy and an adaptive search mechanism, the accuracy and quality of the Pareto solution are effectively improved. This is based on the search progress ratio. The following adaptive search direction adjustment mechanism was designed:

[0079] Search direction in the formula Through historical direction and the current optimal direction The weighted combination was obtained, realizing the cumulative utilization of search experience.

[0080] Based on the above mechanism, the particle position update rule is as follows:

[0081] In the formula, Given the current search radius, exponential decay ensures that the search accuracy gradually improves with each iteration; An adaptive learning rate is used to balance the weights of historical and current information. For the search direction, It is a function that generates random numbers that follow a standard normal distribution (Gaussian distribution).

[0082] The complete population B update strategy is as follows:

[0083] A hierarchical search framework, such as Figure 6 As shown, it specifically includes an L1 layer (wide-area search layer) and an L2 layer (local Pareto fine-grained search layer): Layer L1 serves as a wide-area search layer, covering the entire feasible target space of the problem. Layer L2 serves as a local Pareto fine-grained search layer, limiting its search scope to the high-quality region inherited from Layer L1.

[0084] Next, the population is updated and information is exchanged. The dual-population information exchange mechanism works as follows: during each iteration, both populations write their respective Pareto high-quality solutions into a shared information set, and can simultaneously obtain search information from each other to optimize their own evolutionary direction. To prevent premature convergence due to excessive information exchange, the algorithm employs an adaptive probability control mechanism based on fitness differences.

[0085] In the formula It is the probability of information exchange between populations. This represents the difference in fitness between the current optimal solutions of two populations. and These are adaptive control parameters. When When the probability is large, a lower exchange probability helps maintain the independence and diversity of the population; when When the probability of swapping is small, a higher probability of swapping helps to accelerate the convergence of the algorithm. and The parameter values ​​are dynamically adjusted during the iteration process to adapt to the needs of different search stages. In the bidirectional information exchange mechanism based on Pareto information sharing sets, The layer will pass on the information of the high-quality areas it discovers. Layers provide precise guidance for local search directions; simultaneously The high-quality solution after layer optimization will be fed back to Layer by layer, continuously improve the global Pareto search strategy to achieve efficient integration of global exploration and local development.

[0086] Finally, a convergence check is performed: the algorithm is checked to see if it meets the convergence condition. If the maximum number of iterations is reached or the change in fitness value is less than a preset threshold, the algorithm terminates and outputs the global optimal solution; otherwise, it returns to continue iterating over the two population values.

[0087] Implementation Method 2 This embodiment provides an underwater dual-arm robot target grasping coordination planning system based on dual-population hierarchical particle swarm optimization. The system uses the underwater dual-arm robot target grasping coordination planning method based on dual-population hierarchical particle swarm optimization as described in Embodiment 1. The system includes a controller and its built-in planning and optimization module. The planning and optimization module is used to establish a multi-objective motion planning model for the underwater dual-arm robot and run a dual-population hierarchical particle swarm optimization algorithm to generate coordinated planning results. Based on the coordination planning results, the controller drives the dual-arm robot to autonomously complete target grasping and collaborative operation tasks. The planning results are as follows: Figure 7 , Figure 8 As shown, a coordinated planning method for target grasping of an underwater dual-arm robot based on dual-population hierarchical particle swarm optimization was implemented.

[0088] Implementation Method 3 This invention provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor. The memory stores software programs and modules, and the processor executes various functional applications and data processing by running the software programs and modules stored in the memory. The memory and processor are connected via a bus. Specifically, the processor implements any step in Embodiment 1 by running the computer program stored in the memory.

[0089] It should be understood that, in the embodiments of the present invention, the processor may be a Central Processing Unit (CPU), but it may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0090] Memory may include read-only memory, flash memory, and random access memory, and provides instructions and data to the processor. Some or all of the memory may also include non-volatile random access memory.

[0091] It should be understood that if the integrated modules / units described above are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods described above can also be implemented by a computer program instructing related hardware. This computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content contained in the computer-readable storage medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction.

[0092] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0093] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the above device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this invention. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0094] It should be noted that the methods and detailed examples provided in the above embodiments can be incorporated into the apparatus and devices provided in the embodiments for mutual reference, and will not be repeated here.

[0095] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0096] In the embodiments provided by this invention, it should be understood that the disclosed apparatus / terminal devices and methods can be implemented in other ways. For example, the apparatus / device embodiments described above are merely illustrative. For instance, the division of modules or units described above is merely a logical functional division, and in actual implementation, it can be divided in other ways. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0097] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A coordinated planning method for target grasping of an underwater dual-arm robot based on dual-population hierarchical particle swarm optimization, characterized in that, The method includes the following steps: Step S1: Construct a multi-objective motion planning model for the underwater dual-arm robot based on two levels: local optimization task of symmetrical cooperation between the two robotic arms and global optimization task of division of labor and cooperation between the robot hull and the robotic arms. Clarify the corresponding mechanical safety constraints and system safety constraints. Step S2: Construct a dual-population structure of "two arms - two groups", including exploratory population A and development population B; Step S3: Based on the dual-population structure established in step S2, the corresponding hierarchical search framework is as follows: The L1 layer, as a wide-area search layer, covers the entire solution space to cope with the complexity of high-dimensional multi-objective optimization in two arms. Population A is responsible for discovering candidate solutions that satisfy global safety and stability. The L2 layer, as a local fine-grained search layer, limits the scope to the high-quality region passed by the L1 layer, and the population B discovers candidate solutions that satisfy the multi-task constraints and local accuracy requirements of the two arms. Step S4: The exploration population A and the development population B interact with each other through an information sharing set between the upper and lower layers. During each iteration, the two populations write their respective Pareto high-quality solutions into the sharing set and obtain optimization information from each other to update their own search directions. Step S5: Check whether the iteration in step S4 meets the convergence condition; if the maximum number of iterations is reached or the fitness value change is less than the preset threshold, terminate the iteration and output the global optimal solution; otherwise, return to continue iterating the bipopulation value.

2. The method according to claim 1, characterized in that, The multi-objective motion planning model in step S1 includes a path length objective function, a joint change minimization objective function, and the disturbance of the hull by the manipulator's motion. The path length objective function reflects the total movement distance required for the underwater dual-arm robot to complete its task and the effectiveness of the functional division between the hull and the dual arms. It is defined as the sum of the Euclidean distances of the path point sequence: In the formula, Indicates the first The three-dimensional coordinates of each path point; This represents the total number of path points. Represents the total path length of the entire motion process; Specifically, the objective function for minimizing joint changes is defined as the weighted sum of the changes in the joint angles of the robotic arms: In the formula, Indicates the first The change in the angle of each joint; These are the weighting coefficients for the changes in the angles of each joint; Indicates the stability index of the system; The disturbance to the hull caused by the manipulator's motion is specifically as follows: Considering symmetry, the objective function of the restoring torque of the underwater dual-arm robot during its motion is: In the formula, The total restoring torque generated by the left arm; The total restoring torque generated by the right arm; For the robotic arm in The restoring torque component in the direction, for the left arm is The right arm is .

3. The method according to claim 1, characterized in that, The safety constraints in step S1 include mechanical safety constraints and system safety constraints, specifically as follows: Safety constraints on bi-arm joint angles, velocities, and accelerations, as well as integrated visual field-obstacle avoidance constraints; Risk constraints related to the bottoming out of the arms and the entire hull: In the formula, This indicates the real-time vertical distance between the underwater dual-arm robot and the seabed. , The upper and lower threshold values ​​for this distance are defined respectively.

4. The method according to claim 3, characterized in that, The obstacle avoidance safety constraints and mutual obstacle avoidance constraints between the two arms of the underwater dual-arm robot system are as follows: Using the polygon method, obstacles are enclosed by spheres in three-dimensional space, while the links and hull of the underwater dual-arm robot are simplified to cylinders. The three-dimensional model is converted into a two-dimensional model through mathematical mapping. The underwater dual-arm robot meets the minimum safety distance requirements. : In the formula, The shortest distance from the obstacle to each segment of the robotic arm. This represents the actual minimum distance between any part of an underwater dual-arm robot and an obstacle. For a predefined safe distance threshold; Define the minimum safe distance constraint between the two robotic arms as follows: In the formula, This represents the actual minimum distance between the two robotic arms. The preset safety threshold; When calculating the minimum distance between the two arms, consider the distances between all link pairs: In the formula, For the left arm The first link and the right arm The minimum distance between links.

5. The method according to claim 1, characterized in that, The exploration population A in step S2 specifically undertakes the global discovery task of the Pareto front, focusing on the global optimization of the overall coordination between the hull and the two arms, and is responsible for finding potential non-dominated solution regions in the entire target space, as shown in the equation: in: In the formula, Represents population size, For the first The formula utilizes the position of each particle to fully leverage information from all particles in the population, including information used in the traditional PSO algorithm. Item and Term, while retaining the acceleration coefficient and To maintain the basic characteristics of the algorithm; when When, it indicates that the particle Relative to particles It has a better fitness value; For any two chosen attractors or repulsions and , Its information interaction expression is: In the formula, the adaptive random coefficients , Randomly select values ​​from the interval [-0.05, 0.05]. Indicates the current iteration number. Maximum number of iterations For control parameters; Weighting coefficient The calculation is performed in a normalized form: In the formula, and They represent particles respectively and Fitness value, weight coefficient Then take as This forms an asymmetric weight pair. It is a very small positive number; The proposed velocity update rule leads to an improved particle evolution formula: 。 6. The method according to claim 5, characterized in that, The development population B in step S2 specifically involves, based on the search progress ratio Design an adaptive search direction adjustment mechanism: Search direction in the formula Through historical direction and the current optimal direction The weighted combination is obtained; Based on the above mechanism, the particle position update rule is as follows: In the formula, Given the current search radius, exponential decay ensures that the search accuracy gradually improves with each iteration; Adaptive learning rate; For the search direction, It is a function that generates standard normally distributed random numbers; The complete population B update strategy is as follows: 。 7. The method according to claim 1, characterized in that, The dual-population information interaction mechanism in step S4 is as follows: In each iteration, the two populations write their respective Pareto high-quality solutions into the information sharing set, and can also obtain search information provided by each other to optimize their own evolutionary direction; an adaptive probability control mechanism based on fitness differences is adopted. In the formula, It is the probability of information exchange between populations. This represents the difference in fitness between the current optimal solutions of two populations. and These are adaptive control parameters; and The parameter values ​​will be dynamically adjusted during the iteration process.

8. A coordinated planning system for target grasping of an underwater dual-arm robot based on dual-population hierarchical particle swarm optimization, characterized in that, The system uses the underwater dual-arm robot target grasping coordination planning method based on dual-population hierarchical particle swarm optimization as described in any one of claims 1-7, and the system includes a controller and its built-in planning and optimization module. The planning and optimization module is used to establish a multi-objective motion planning model for the underwater dual-arm robot and run a dual-population hierarchical particle swarm optimization algorithm to generate coordinated planning results. Based on the coordination planning results, the controller drives the dual-arm robot to autonomously complete target grasping and collaborative operation tasks.

9. A computer device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1-7.

Images