A redundant robot inverse kinematics parameter solving method and system based on a chaotic botryllus schlosseri algorithm

Abstract

The application belongs to the technical field of robot inverse kinematics solving, and discloses a redundant robot inverse kinematics parameter solving method and system based on a chaos sea slug algorithm, which comprises the following steps: obtaining a relationship expression between a robot end effector position and direction and inverse kinematics parameters to be solved; constructing a fitness function; setting parameters required by the chaos sea slug algorithm and initializing a population based on a Logistic chaos mapping; calculating the fitness F of all individuals in the initial population by using the fitness function, taking the individual with the minimum fitness value as a food source; taking the 1+N / 2 individuals with the highest fitness values as leaders, and taking the remaining individuals as followers, wherein N is the population size; taking the food source as a target, updating the positions of the leaders by using a TLC chaos mapping method, and updating the positions of the followers by using a dynamic learning method until the iteration ends to obtain the robot inverse motion parameters. The application realizes fast and accurate solving of robot inverse motion.

Description

Technical Field

[0001] This invention belongs to the technical field of robot inverse kinematics solution, and more specifically, relates to a method and system for solving redundant robot inverse kinematics parameters based on the chaotic tunic algorithm. Background Technology

[0002] Inverse kinematics plays a crucial role in robotics, involving the calculation of joint angles required to achieve a target position and orientation of the robot's end effector. Common solution methods include geometric, analytical, and iterative methods. Geometric methods establish a set of position constraint equations based on the geometric relationships of the robot mechanism. While computationally simple, they are only applicable to specific mechanisms. Analytical methods utilize specific robot mechanisms to establish analytical formulas for the end effector. Their advantage lies in simplicity and efficiency, but they are only suitable for robots with simple structures, such as a 6-DOF robotic arm satisfying the Piper criterion. These methods are often ineffective in calculating the inverse kinematics of a nine-DOF redundant robot like CMOR. Iterative methods solve the inverse kinematics problem step by step through iterative approximation, have a wide range of applications, but are sensitive to initial values, require a large amount of computation, and have slow convergence speeds. These traditional methods are limited by configuration constraints and are unsuitable for solving the inverse kinematics of the nine-DOF redundant robot CMOR.

[0003] With the emergence of intelligent optimization methods, researchers have defined the inverse kinematics problem as an optimization problem, seeking the optimal combination of joint angles. These methods include Particle Swarm Optimization (PSO), Firefly Algorithm (FA), Differential Evolution (DE), Simulated Annealing (SA), and Squirting Alias ​​Algorithm (SSA). These intelligent optimization methods are applicable to various types and configurations of robots, including those with highly redundant and complex configurations. However, these existing methods face challenges in computational efficiency and long computation times when dealing with complex robot configurations and unique workspace constraints. Especially for complex and redundant robots like CMOR, traditional intelligent optimization methods may not be able to effectively solve their inverse kinematics problem. Summary of the Invention

[0004] To address the aforementioned deficiencies or improvement needs of existing technologies, this invention provides a method for solving the inverse kinematics parameters of redundant robots based on the chaotic tunic algorithm. Its purpose is to achieve a fast and accurate solution for the inverse motion of robots that do not satisfy the Piper criterion.

[0005] To achieve the above objectives, according to one aspect of the present invention, a method for solving redundant robot inverse kinematics parameters based on the chaotic tunicate algorithm is provided, comprising: S1: obtaining the relational expression between the position and orientation of the robot's end effector and the inverse kinematics parameters to be solved; S2: defining the comprehensive error value between the actual pose and the target pose of the end effector as a fitness function; S3: using the robot's inverse kinematics parameters as individuals in the population, setting the parameters required by the chaotic tunicate algorithm and initializing the population based on the Logistic chaotic mapping; S4: calculating the fitness F of all individuals in the initial population using the fitness function. The fitness values ​​are sorted in descending order, and the individual with the smallest fitness value is selected as the food source in the current population; S5: The top 1+N / 2 individuals with the highest fitness values ​​are selected as leaders, and the remaining N / 2-1 individuals are selected as followers, where N is the population size; S6: Using the food source as the target, the Tent-Logistic-Cosine chaotic mapping method is used to update the leader position, and the dynamic learning method is used to update the follower position until the iteration ends and the optimal population individual position is obtained. The optimal population individual position is then substituted into the relational expression to obtain the optimal population individual, which is also the robot's inverse motion parameter.

[0006] Preferably, step S1 specifically includes: S11: defining the DH parameters of each joint according to the Denavit-Hartenberg parameter method and establishing a coordinate system for the link to obtain the homogeneous transformation matrix of each joint; S12: obtaining the forward kinematic equation of the robot by multiplying the homogeneous transformation matrix of each joint.

[0007] Preferably, step S11 specifically includes: establishing a coordinate system for each joint and link according to the Denavit-Hartenberg parameter method; and constructing a homogeneous transformation matrix for each joint based on the DH parameters of the joint and link in the coordinate system.

[0008] Preferably, the fitness function is:

[0009] F = E p +λ·E o

[0010] Where F is the fitness function, E p E represents the error between the actual position and the target position of the robot's end effector. o Let λ be the error value between the actual posture and the target posture of the robot's end effector, and λ be the adjustment coefficient.

[0011] Preferably, the population initialization equation using the Logistic chaotic mapping is:

[0012] x i,j =Lbj +Cl i,j (Ub j -Lb j )

[0013] Where i represents the population size, i = 1, 2, ..., N, j represents the chaos index, j = 1, 2, ..., D, and D represents the population dimension, which is also the number of parameters that need to be optimized in the inverse solution. i,j Let [Lb] be the j-dimensional position of the i-th individual in the population. j Ub j [ ] represents the range of motion of the j-th joint of the robot, Cl i,j Let be the chaotic value corresponding to the j-dimensional position of the i-th individual within the Logistic chaotic sequence.

[0014] Preferably, for the leaders, the Tent-Logistic-Cosine chaotic mapping method is used to update the position of each leader using the following specific formula:

[0015]

[0016] Where i = 1, 2, ..., L, C TLC To represent the chaotic values ​​of a chaotic sequence generated using the Tent-Logistic-Cosine chaotic map, x i,j Let Fd be the j-dimensional position of the i-th leader. j Let c be the j-dimensional location of the food source, c3 be a random number in the interval [0,1], and c1 be the convergence factor. t is the current iteration number, and T is the maximum iteration number.

[0017] Preferably, the follower position update formula used in step S6 is as follows:

[0018] x i,j =ω t ·Fd j +(x i-1,j +x i,j ) / 2

[0019] Where i = L+1, L+2, ..., N, x i,j Let Fd be the j-dimensional position of the i-th follower. j Let ω be the j-dimensional position of the food source. t For learning weights, ω t =ω max -(ω max -ω min )·T / t, where t is the current iteration number, T is the maximum iteration number, and ω max ω is the maximum value of the learned weights. minThis represents the minimum value of the learning weight.

[0020] Preferably, step S6 includes: if the change after a preset number of iterations is less than a preset threshold, then a random perturbation is added to the current location of the food source, and the perturbed location of the food source is:

[0021] Fd j =Fd j +δ·Cd

[0022] Where Cd is a random number in the range [0,1].

[0023] Preferably, δ is 0.01% to 1%.

[0024] The second aspect of this application provides a system for solving the inverse kinematics parameters of a redundant robot based on the chaotic tunicate algorithm, comprising: an acquisition module for acquiring the relationship expression between the position and orientation of the robot's end effector and the inverse kinematics parameters to be solved; a fitness function determination module for defining the comprehensive error value between the actual pose and the target pose of the robot's end effector as the fitness function; an initialization module for setting the parameters required by the chaotic tunicate algorithm and initializing the population based on the Logistic chaotic mapping, using the robot's inverse kinematics parameters as individuals in the population; and a food source determination module for calculating the fitness F of all individuals in the initial population using the fitness function. The fitness values ​​are sorted in descending order, and the individual with the smallest fitness value is selected as the food source in the current population. The population role determination module identifies the top 1+N / 2 individuals as leaders and the remaining N / 2-1 individuals as followers, where N is the population size. The iterative solution module uses the Tent-Logistic-Cosine chaotic mapping method to update the leader positions and a dynamic learning method to update the follower positions, with the food source as the target, until the iteration ends and the optimal population individual position is obtained. This optimal population individual position is then substituted into the relational expression to obtain the optimal population individual, which is also the robot's inverse motion parameter.

[0025] In summary, compared with the prior art, the method and system for solving inverse kinematics parameters of redundant robots based on the chaotic tunic algorithm provided by this invention have the following advantages:

[0026] 1. This application addresses the design of redundant serial robots that do not satisfy the Piper criterion. It aims to transform the complex inverse kinematics problem into an optimization problem of solving a system of nonlinear equations, utilizing the chaotic jar-and-squirt algorithm to find a set of joint variables (θ1, θ2, ..., θ...). n This ensures that the robot's end effector can accurately reach the expected position and orientation.

[0027] 2. The chaotic tunicate algorithm employed in this application is not limited by the initial robot configuration during its iterative process. By introducing chaos theory to perform chaotic initialization of the population, the algorithm's search capability and global exploration performance are significantly improved. Compared to traditional optimization algorithms, this method exhibits higher efficiency and better solution quality when dealing with inverse kinematics problems under high-dimensional and complex constraints. Furthermore, the algorithm possesses good scalability and versatility, making it suitable for various types of redundant serial robot inverse kinematics problems.

[0028] 3. In the standard tunic algorithm, the leader is a single individual at the head of the population chain, with the remaining N-1 individuals serving as followers. However, this extremely small number of leaders is severely insufficient during the early stages of the algorithm's iterations for global exploration, leading to inadequate global exploration, overly refined local development, and a tendency to get trapped in local optima. This application improves the selection of leaders and followers to avoid getting trapped in local optima.

[0029] 4. The Tent-Logistic-Cosine (TLC) chaotic mapping is adopted. The TLC chaotic mapping generates pseudo-random distribution state values ​​in each iteration, which expands the randomness and ergodicity of the SSA search process, thereby achieving a more reasonable search process than the completely uniform distribution.

[0030] 5. During the iteration process, if the algorithm is found to be trapped in a local optimum, this application adds a random perturbation to the position of the current solution to avoid the algorithm from getting trapped in a local optimum. Attached Figure Description

[0031] Figure 1 This is a schematic diagram illustrating the steps of the method for solving the inverse kinematics parameters of a redundant robot based on the chaotic tunic algorithm in this application embodiment;

[0032] Figure 2 This is a flowchart illustrating the method for solving the inverse kinematics parameters of a redundant robot based on the chaotic tunic algorithm in this application embodiment;

[0033] Figure 3 This is a schematic diagram of the DH coordinate system of the robot with the ninth degree of freedom in Embodiment 9 of this application. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0035] The first aspect of this application provides a method for solving the inverse kinematics parameters of a redundant robot based on the chaotic tunic algorithm, such as... Figure 1 and Figure 2 As shown, the process includes the following steps S1 to S6.

[0036] S1: Obtain the relationship expression between the position and orientation of the robot's end effector and the inverse kinematic parameters to be solved.

[0037] Step S1 specifically includes S11 and S12:

[0038] S11: Define the DH parameters for each joint according to the DH parameter method and establish a coordinate system for the link to obtain the homogeneous transformation matrix for each joint.

[0039] A coordinate system is established for each joint and link using the Denavit-Hartenberg (DH) parametric method. The DH parameters include four parameters: joint angle θ, joint offset d, link length a, and link torsion angle α.

[0040] In this embodiment, the structure of the nine-degree-of-freedom redundant robot CMOR (such as...) Figure 3 Taking the example shown in Table 1, the DH parameters of each joint are determined to form a parameter table.

[0041] Table 1

[0042] 1 0 0 90 0-5m 2 <![CDATA[2]]> 0 a2 0 ±70° 3 <![CDATA[3+90]]> 0 0 90 ±85° 4 <![CDATA[4-90]]> <![CDATA[d4]]> a4 -90 ±180° 5 <![CDATA[5+180]]> 0 a5 -90 0-180 6 <![CDATA[6-90]]> <![CDATA[d6]]> 0 90 ±180° 7 <![CDATA[7+180]]> 0 0 90 ±90° 8 <![CDATA[8+180]]> <![CDATA[d8]]> 0 90 ±180° 9 <![CDATA[9+180]]> 0 0 0 ±90°

[0043] The transformation matrix of coordinate system {i} relative to coordinate system {i-1} is established using the DH parameters. This matrix is ​​a 4×4 homogeneous transformation matrix. Specifically, for the i-th joint, its homogeneous transformation matrix T... i i-1 It can be represented as:

[0044]

[0045] Based on the above transformation matrices and Table 1, the transformation matrix for each link can be obtained:

[0046]

[0047]

[0048] Here, s and c are abbreviations for sin and cos functions.

[0049] S12: Based on the homogeneous transformation matrix of each joint, the forward kinematic equation of the robot is obtained by multiplication.

[0050] By multiplying the homogeneous transformation matrix of each joint step by step, the total transformation matrix of the end effector relative to the base coordinate system is obtained.

[0051]

[0052] Where n is the number of joints in the robot.

[0053] This transformation matrix describes the position and orientation of the end effector in three-dimensional space:

[0054]

[0055] Therefore, the robot's forward kinematics equations can also be expressed as:

[0056] F forward (θ1,θ2,…,θ n )=(n 3×1 ,o 3×1 ,a 3×1 ,p x ,p y ,p z )

[0057] S2: Define the combined error value between the actual pose and the target pose of the robot end effector as the fitness function.

[0058] The inverse kinematics problem of the robot is transformed into an optimization problem. The fitness function F is defined as the error value between the actual position and orientation of the robot's end effector and the target position and orientation, expressed as:

[0059]

[0060] F = E p +λ·E o

[0061] Where F is the fitness function, E p E represents the error between the actual position and the target position of the robot's end effector. o This represents the error value between the actual posture and the target posture of the robot's end effector.

[0062] Since position and attitude errors have different dimensions, position is usually quantified in units of length (such as meters), while attitude is quantified in angles (such as radians). A coefficient needs to be added to appropriately scale these two different types of errors so that they are comparable in the optimization process. The preferred coefficient is λ, where λ = 0.015.

[0063] S3: Using the robot's inverse kinematics parameters as the population individuals, set the parameters required for the chaotic tunic algorithm and perform population initialization based on the Logistic chaotic mapping.

[0064] In this embodiment, the robot's inverse kinematics parameters are joint angles. Each individual in the population represents a potential solution, and therefore each individual represents a specific set of joint angles. These parameters define a specific posture of the robot.

[0065] The Modified Slap Swarm Algorithm based on chaotic maps (MCSSA) requires the following parameters: population size N, dimension D, maximum number of iterations T, convergence factor c1, random parameters c3 and Cd, and initial chaotic values ​​Cl(0) and Cd. TLC (0), Location boundary [Lb] j Ub j Learning weight range ω max and ω min , Preset threshold ε.

[0066] Population initialization based on Logistic chaotic mapping is as follows:

[0067] In D-dimensional space, the initial population is:

[0068] X i =(x i，1 ,x i，2 ,x i，3 ,...,x i，D ), i = 1, 2, ..., N

[0069] Where N is the population size, X i For individuals in a population of salps, each individual typically represents a potential solution.

[0070] To increase the diversity of the initial population distribution and thus improve the search performance of the algorithm, this application utilizes the inherent randomness and complexity of chaotic motion to perform chaotic initialization of the population, which can effectively explore different regions of the solution space. A chaotic sequence is generated using a Logistic map.

[0071] Cl i,j+1 =4Cl i,j (1-Cl i,j )

[0072] Among them, Cl i,j Let Cl(0) be the chaotic value generated in the interval [0,1], and let Cl(0) be the initial value of the iteration, which is randomly generated.

[0073] Therefore, the population initialization process of the chaotic tunic algorithm is as follows:

[0074] x i,j =Lb j +Cli,j (Ub j -Lb j )

[0075] Where i represents the population size, i = 1, 2, ..., N, j represents the chaos index, j = 1, 2, ..., D, and D represents the population dimension, which is also the number of parameters that need to be optimized in the inverse solution. i,j Let [Lb] be the j-dimensional position of the i-th individual in the population. j Ub j [ ] represents the search range of the population, and Cl represents the range of motion of the j-th joint of the robot. i,j Let be the chaotic value corresponding to the j-dimensional position of the i-th individual within the Logistic chaotic sequence.

[0076] S4: Calculate the fitness F of all individuals in the initial population using the fitness function, sort the fitness values ​​in descending order, and select the individual with the smallest fitness value as the food source in the current population.

[0077] The fitness value F of all individuals N in the initial salps population is calculated based on the fitness function F. Based on the fitness value of the individuals in the initial salps population, all salps individuals are sorted in descending order, and the salps individual with the smallest fitness value is taken as the food source position Fd in the current population.

[0078] S5: Select the top 1+N / 2 individuals by fitness value as leaders, and the remaining N / 2-1 individuals as followers, where N is the population size.

[0079] In the standard Squirting Aquarius (SSA) algorithm, the population size is N, where the leader is the individual at the head of the population chain, numbering 1, and the remaining N-1 individuals are followers. However, this extremely small number of leaders is severely insufficient during the early stages of the algorithm's iterations when conducting global exploration, leading to inadequate global exploration, overly refined local development, and a tendency to get trapped in local optima.

[0080] To overcome the above limitations, this invention proposes to designate the top 1+N / 2 individuals in the fitness ranking of the salps population as leaders, and the remaining N / 2-1 individuals as followers.

[0081] S6: Taking the food source as the target, the leader position is updated using the Tent-Logistic-Cosine chaotic mapping method, and the follower position is updated using the dynamic learning method. During the iteration process, local optima are monitored and randomly perturbed until the iteration ends and the optimal population individual position is obtained. The optimal population individual position is then substituted into the relational expression to obtain the optimal population individual, which is also the robot's inverse motion parameter.

[0082] In the aforementioned new population structure, the leader performs a global search targeting food sources. The leader's position update is mainly influenced by two factors: a convergence factor c1 and a scaling factor c2. c1 varies with the number of iterations, while c2 is a uniformly distributed random parameter that determines the algorithm's global exploration range. To improve the effectiveness of the global search and reduce blindness, this application introduces the Tent-Logistic-Cosine (TLC) chaotic mapping. The TLC chaotic mapping generates pseudo-random distributed state values ​​in each iteration, extending the randomness and ergodicity of the search process of the tunic algorithm, thus achieving a more reasonable search process than a completely uniform distribution. Compared to the more complex high-dimensional chaotic mapping, the TLC mapping is easier to implement due to its relatively simple structure. The mathematical formula for the TLC mapping is shown below:

[0083]

[0084] Where r is the parameter of the TLC chaotic mapping, which is set here to r = 0.7, and the initial value of the iteration is C. TLC (0) is randomly generated, C TLC (t+1) represents the chaotic value generated by the TLC chaotic mapping in iteration t.

[0085] For the leaders (the number of individuals is L = 1 + N / 2), the method for updating the position of each leader based on the TLC chaotic mapping is as follows:

[0086]

[0087] Where i = 1, 2, ..., L, C TLC To represent the chaotic values ​​of a chaotic sequence generated using the Tent-Logistic-Cosine chaotic map, x i,j Let Fd be the j-dimensional position of the i-th leader. j Let c be the j-dimensional location of the food source, c3 be a random number in the interval [0,1], and c1 be the convergence factor, defined as c1 = 2e -(4t / T)2 t is the current iteration number, and T is the maximum iteration number.

[0088] Furthermore, in the Standard Squirting Algorithm (SSA), followers are primarily responsible for the local search process. The follower's position update is unaffected by any random parameters, relying solely on the positions of neighboring individuals, without considering the leader's position. This mechanism ensures sufficient local search capability, but also means that if the leader gets trapped in a local optimum, the follower will struggle to escape. To overcome this limitation, this application borrows from the social behavior component of particles in the Particle Swarm Optimization (PSO) algorithm, enhancing the influence of elite individuals (i.e., leaders) on followers, and introduces a follower position update mechanism based on dynamically learned adaptive inertia weights. For followers (with the number of individuals F = N / 2 - 1), the specific update formula is:

[0089] x i,j =ω t ·Fd j +(x i-1,j +x i,j ) / 2

[0090] Where i = L+1, L+2, ..., N, x i,j Let Fd be the j-dimensional position of the i-th follower. j Let ω be the j-dimensional position of the food source. t For learning weights, ω t =ω max -(ω max -ω min )·T / t, where t is the current iteration number, T is the maximum iteration number, and ω max ω is the maximum value of the learned weights. min To minimize the learning weights, in this embodiment, ω max =0.9,ω min =0.2.

[0091] To avoid the algorithm getting trapped in local optima, monitoring the iterative process of the solution is essential. Once the algorithm gets trapped in a local optimum, a random perturbation is added to the current food source position. The magnitude of the random perturbation typically depends on the dimension of the solution and the desired perturbation strength, generally between 0.01% and 1%. Considering that this algorithm is applied to the joint angles of a robot, and in radians, the introduced random perturbation is relatively small.

[0092] If the change after a preset number of iterations is less than a preset threshold, for example, if the change after more than 25 iterations is less than the preset threshold of 0.01%, then the algorithm is considered to have entered a local optimum. In this case, a random perturbation is added to the current food source location, and the perturbed food source location is:

[0093] Fd j =Fd j +δ·Cd

[0094] Where Cd is a random number in the range [0,1], and δ is set to 0.01%.

[0095] The evaluation and update process is repeated until the stopping condition (maximum number of iterations) is met. During this process, each individual instance of the chaotic tunic algorithm continuously approaches the optimal solution. Ultimately, the algorithm produces a set of optimal or near-optimal joint angles, which represent the solution to the robot's inverse kinematics.

[0096] The second aspect of this application provides a system for solving the inverse kinematics parameters of a redundant robot based on the chaotic tunic algorithm, the system comprising:

[0097] Acquisition module: used to acquire the relationship expression between the position and orientation of the robot's end effector and the inverse kinematics parameters to be solved;

[0098] Fitness function determination module: used to define the comprehensive error value between the actual pose and the target pose of the robot end effector as the fitness function;

[0099] Initialization module: Used to set the parameters required by the chaotic tunicate algorithm and perform population initialization based on the Logistic chaotic mapping, using the robot's inverse kinematics parameters as the population individuals;

[0100] Food source determination module: used to calculate the fitness F of all individuals in the initial population using the fitness function, sort the fitness values ​​in descending order, and select the individual with the smallest fitness value as the food source in the current population;

[0101] Population role determination module: used to designate the top 1+N / 2 individuals by fitness value as leaders, and the remaining N / 2-1 individuals as followers, where N is the population size;

[0102] Iterative solution module: It is used to update the leader position using the Tent-Logistic-Cosine chaotic mapping method and the follower position using the dynamic learning method, with the food source as the target, until the iteration ends and the optimal population individual position is obtained. The optimal population individual position is then substituted into the relational expression to obtain the optimal population individual, which is also the robot's inverse motion parameters.

[0103] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for solving the inverse kinematics parameters of a redundant robot based on the chaotic tunic algorithm, characterized in that, include: S1: Obtain the relationship expression between the position and orientation of the robot's end effector and the inverse kinematics parameters to be solved; S11: Define the DH parameters of each joint according to the Denavit-Hartenberg parameter method and establish a coordinate system for the link to obtain the homogeneous transformation matrix of each joint; Step S11 specifically includes: Establish a coordinate system for each joint and link according to the Denavit-Hartenberg parametric method; In the coordinate system, construct the homogeneous transformation matrix of each joint based on the DH parameters of the joint and the link; S12: Based on the homogeneous transformation matrix of each joint, the forward kinematic equation of the robot is obtained by multiplication. S2: Define the combined error value between the actual pose and the target pose of the robot end effector as the fitness function; S3: Using the robot's inverse kinematics parameters as the population individuals, set the parameters required by the chaotic tunicate algorithm and perform population initialization based on the Logistic chaotic mapping; S4: Calculate the fitness F of all individuals in the initial population using the fitness function, sort the fitness values ​​in descending order, and select the individual with the smallest fitness value as the food source in the current population; S5: Select the top 1+N / 2 individuals by fitness value as leaders, and the remaining N / 2-1 individuals as followers, where N is the population size; S6: Taking the food source as the target, the leader position is updated using the Tent-Logistic-Cosine chaotic mapping method, and the follower position is updated using the dynamic learning method until the iteration ends and the optimal population individual position is obtained. The optimal population individual position is then substituted into the relational expression to obtain the optimal population individual, which is also the robot's inverse motion parameter.

2. The method for solving the inverse kinematics parameters of a redundant robot according to claim 1, characterized in that, The fitness function is: Where F is the fitness function, This represents the error value between the actual position and the target position of the robot's end effector. This represents the error value between the actual posture and the target posture of the robot's end effector. This is for adjusting the coefficient.

3. The method for solving the inverse kinematics parameters of a redundant robot according to claim 1, characterized in that, The population initialization equation using the Logistic chaotic mapping is as follows: in, i Indicates population size, i =1,2,…,N, j Indicates the chaotic sequence number. j =1,2,…,D, where D represents the dimension of the population, which is also the number of parameters that need to be optimized in the inverse solution; The first in the population i individual j Dimensional position, For the robot j Range of motion of each joint For the first i individual j The chaotic value corresponding to the dimensional position within the Logistic chaotic sequence.

4. The method for solving the inverse kinematics parameters of a redundant robot according to claim 3, characterized in that, For the leaders, the Tent-Logistic-Cosine chaotic mapping method is used to update the position of each leader using the following specific formula: Where i = 1, 2, ..., L, To represent the chaotic values ​​of a chaotic sequence generated using the Tent-Logistic-Cosine chaotic map. For the first i The j-dimensional position of a leader Let j be the j-dimensional location of the food source. A random number within the interval [0,1]. Let be the convergence factor, defined as t is the current iteration number, and T is the maximum iteration number.

5. The method for solving the inverse kinematics parameters of a redundant robot according to claim 1, characterized in that, In step S6, for followers, a dynamic learning method is used to update the follower positions. The specific update formula is as follows: in, i =L+1,L+2,…,N, For the first i The j-dimensional position of each follower Let j be the j-dimensional location of the food source. For learning weights, t is the current iteration number, and T is the maximum iteration number. To be the maximum value of the learning weights, This represents the minimum value of the learning weight.

6. The method for solving the inverse kinematics parameters of a redundant robot according to claim 1, characterized in that, Step S6 includes: if the change after a preset number of iterations is less than a preset threshold, then a random perturbation is added to the current location of the food source, and the perturbed location of the food source is: in, A random number in the range [0,1]. It ranges from 0.01% to 1%.

7. A system for solving inverse kinematics parameters of a redundant robot based on the chaotic tunic algorithm, which incorporates the method described in any one of claims 1-6, characterized in that, include: Acquisition module: used to acquire the relationship expression between the position and orientation of the robot's end effector and the inverse kinematics parameters to be solved; Fitness function determination module: Defines the comprehensive error value between the actual pose and the target pose of the robot end effector as the fitness function; Initialization module: Used to set the parameters required by the chaotic tunicate algorithm and perform population initialization based on the Logistic chaotic mapping, using the robot's inverse kinematics parameters as the population individuals; Food source determination module: used to calculate the fitness F of all individuals in the initial population using the fitness function, sort the fitness values ​​in descending order, and select the individual with the smallest fitness value as the food source in the current population; Population role determination module: used to designate the top 1+N / 2 individuals by fitness value as leaders, and the remaining N / 2-1 individuals as followers, where N is the population size; Iterative solution module: It is used to update the leader position using the Tent-Logistic-Cosine chaotic mapping method with food source as the target, and to update the follower position using the dynamic learning method until the iteration ends and the optimal population individual position is obtained. The optimal population individual position is then substituted into the relational expression to obtain the optimal population individual, which is also the robot inverse motion parameter.

Images