Control method, system and equipment of multi-mechanical-arm system and medium

Abstract

The invention discloses a control method, system and equipment of a multi-mechanical-arm system and a medium, and relates to the field of multi-mechanical-arm system.The control method comprises the steps that in a nonlinear system, a composite disturbance observer is used for estimating the composite disturbance state of each mechanical arm; a local state observer and a distributed observer are used for each mechanical arm, and a local state observation error and a distributed observation error of each mechanical arm are obtained; determining the consistent bounded state of each mechanical arm according to the local state observation error and the distributed observation error; and according to the composite disturbance state and the consistent bounded state, the nonlinear system keeps stable operation, and consistent tracking control of the nonlinear system can be realized. Accurate control over the multi-mechanical-arm system can be achieved.

Description

Technical Field

[0001] The present invention relates to the field of multi-robot system control, and in particular to a control method, system, equipment and medium for a multi-robot system. Background Art

[0002] With the rise of robotics and artificial intelligence technology, robotic arms play an important role in industrial production, modern medicine, space exploration, etc. In the field of multi-robotic arm system control, traditional control methods mainly include PID control, sliding mode control, fuzzy logic control and model-based predictive control.

[0003] In the prior art, the first method uses sliding mode control, which is favored for its good robustness and tolerance to parameter uncertainty. It achieves the control purpose by designing a sliding surface so that the system state converges quickly to the surface and slides along the surface. The second method uses a solution based on model predictive control, which predicts the future state according to the system model and optimizes the control sequence to minimize the cost function. It can handle constraints and is easy to expand to multi-input and multi-output systems.

[0004] However, sliding mode control and model-based predictive control methods lack anti-interference and collaborative operation capabilities in the control of multi-robotic arm systems, resulting in insufficient accuracy and inability to achieve precise control of multi-robotic arm systems. Summary of the invention

[0005] Embodiments of the present invention provide a control method, system, device and medium for a multi-robotic arm system, which can solve the problem in the prior art that accurate control of the multi-robotic arm system cannot be achieved.

[0006] An embodiment of the present invention provides a control method for a multi-robotic arm system, comprising the following steps: constructing a nonlinear system in which a leader robotic arm leads multiple follower robotic arms to move through information interaction; in the nonlinear system, using a composite disturbance observer for each robotic arm to estimate the composite disturbance state of each robotic arm; using a local state observer and a distributed observer for each robotic arm respectively to obtain a local state observation error and a distributed observation error of each robotic arm; determining a consistent bounded state of each robotic arm according to the local state observation error and the distributed observation error; keeping the nonlinear system in stable operation according to the composite disturbance state and the consistent bounded state, and realizing consistent tracking control of the nonlinear system, wherein the consistent tracking control is used to indicate that all robotic arms can achieve consistent motion.

[0007] Furthermore, the obtaining of the local state observation error and the distributed observation error of each robot arm specifically comprises the following steps:

[0008] Use a local state observer for each robot

[0009]

[0010] The local state observation error is defined as

[0011] Where N represents the number of follower robotic arms, i represents the i-th follower robotic arm, is the state x i The estimate, is the output of the state observer, u i ∈R p is the control input, K∈R n×q is the gain matrix, making A+KC a Hurwitz matrix; θ i (t) represents the input delay and satisfies θ i (t) = t - τ i (t), τ i (t) is the non-uniform time-varying input delay; represents uncertain external disturbance; and are the weight matrix and activation function of the radial basis neural network, A∈R n×n , B∈R n×p , C∈R q×n is a constant matrix;

[0012] Use a distributed observer for each robot

[0013]

[0014] in, represents an estimate of the leader's state, represents the output of the distributed observer, Λ∈R q×n is the gain matrix, e i is the consistent tracking error, and the formula is:

[0015]

[0016] The distributed observation error is defined as

[0017] Among them, x0 is the leader state, j represents the index of agent j; N i represents the neighboring agent of agent i; a ij represents the communication weight between agent i and agent j; b i represents the communication weight between agent i and the leader; y0 represents the leader output.

[0018] Furthermore, the method of determining the consistent bounded state of each robotic arm according to the local state observation error and the distributed observation error specifically comprises the following steps:

[0019] Get local state observation error and distributed observation errors

[0020] According to the two errors

[0021] When time t tends to infinity, we get and Determine the consistent and bounded state of each robot.

[0022] Furthermore, the implementation of the consistency tracking control of the nonlinear system comprises the following specific steps:

[0023] The consistency tracking control problem of nonlinear system is transformed into the stability analysis of nonlinear system through composite disturbance observer, local state observer and distributed observer, and the stability model of nonlinear system is obtained.

[0024] The stability model of the nonlinear system is as follows:

[0025]

[0026] in, Uncertain external disturbances; Local state observation error; is the estimation error, and They are the weight matrix and activation function of the radial basis neural network, Represents the estimate of the leader state; A∈R n×n , B∈R n×p , C∈R q×n is a constant matrix, K∈R n×q ,Λ∈R q×n is the gain matrix; e i is the consistent tracking error; u i ∈R p is the control input; θ i Indicates input delay; Error information, Follower state state estimation, represents the leader state estimate;

[0027] Constructing a controller according to the stability model of the nonlinear system, wherein the controller can realize a consistency tracking control model of the nonlinear system;

[0028] The controller formula is:

[0029]

[0030] Among them, u i ∈R p is the control input; Γ i =-B T P i , P i is a positive definite matrix; e i is the consistent tracking error; A∈R n×n is a constant matrix; The input delay boundary is a known constant; Error information, Follower state estimation, Leader state estimation.

[0031] An embodiment of the present invention provides a control system for a multi-manipulator system, including:

[0032] A system construction module is used to construct a nonlinear system in which a leader robot arm leads multiple follower robot arms to move through information interaction; a control design module is used to estimate the composite disturbance state of each robot arm using a composite disturbance observer for each robot arm in the nonlinear system; a local state observer and a distributed observer are used for each robot arm to obtain the local state observation error and the distributed observation error; a consistent bounded state of each robot arm is determined based on the local state observation error and the distributed observation error; a control system construction module is used to keep the nonlinear system running stably based on the composite disturbance state and the consistent bounded state, and to achieve consistent tracking control of the nonlinear system, wherein the consistent tracking control is used to indicate that all robot arms can achieve consistent motion.

[0033] An embodiment of the present invention provides a computer device, comprising: a memory and a processor; the memory stores a computer program, and the processor implements the above-mentioned control method of a multi-robotic arm system when executing the computer program.

[0034] An embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the control method of a multi-robotic arm system described above is implemented.

[0035] The embodiments of the present invention provide a control method, system, device and medium for a multi-robot system. Compared with the prior art, the control method, system, device and medium have the following beneficial effects:

[0036] Composite disturbance observers, local state observers and distributed observers are collectively referred to as adaptive observers. Adaptive observers can monitor and compensate in real time the control errors caused by external interference during the controlled operation of the robot arm, avoid drastic fluctuations in the control signal, and ultimately achieve precise control of the multi-robot system. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A schematic diagram of a consistency tracking control scheme provided by an embodiment of the present invention;

[0038] Figure 2 A directed communication topology diagram provided by an embodiment of the present invention;

[0039] Figure 3 A schematic diagram of the local state observation error of a follower manipulator provided in an embodiment of the present invention;

[0040] Figure 4 A schematic diagram of the local state observation error of the leader manipulator provided by an embodiment of the present invention;

[0041] Figure 5 A schematic diagram of system consistency tracking error provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0042] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention, so the present invention is not limited by the specific embodiments disclosed below.

[0043] See also Figure 1 to Figure 5 , an embodiment of the present invention provides a control method for a multi-manipulator system, comprising the following steps:

[0044] Step 1: Construct a nonlinear system in which a leader robot arm leads multiple follower robot arms to move through information interaction.

[0045] Step 2: In the nonlinear system, a composite disturbance observer is used to estimate the composite disturbance state of each robot arm; a local state observer and a distributed observer are used for each robot arm to obtain the local state observation error and the distributed observation error of each robot arm; and the consistently bounded state of each robot arm is determined based on the local state observation error and the distributed observation error.

[0046] Step 3: Based on the composite disturbance state and the consistent bounded state, the nonlinear system can maintain stable operation and achieve consistent tracking control of the nonlinear system. Consistent tracking control is used to indicate that all robotic arms can achieve consistent motion.

[0047] The technical solution of the present invention revolves around a multi-manipulator tracking control method based on an adaptive observer, aiming to improve the tracking accuracy and robustness of the manipulator in a dynamic and uncertain environment.

[0048] The present invention aims to solve three key defects of sliding mode control and model predictive control schemes in the field of multi-manipulator system control, which are as follows:

[0049] 1. Eliminate the "chattering" effect and improve control accuracy

[0050] In sliding mode control, the robot arm may experience "jittering", which is the vibration caused by the high-frequency switching of the control signal, affecting the accuracy and long-term reliability of the robot arm. The present invention, through the introduction of an adaptive observer, can monitor and compensate for the control error caused by parameter changes and external interference in real time, avoiding the drastic fluctuation of the control signal, thereby significantly reducing the "jittering" phenomenon and improving the tracking accuracy and stability of the robot arm.

[0051] 2. Enhance collaborative work capabilities and achieve smooth transition

[0052] Traditional sliding mode control and model predictive control are difficult to ensure smoothness and coordination of movements between multiple robotic arms when working together, especially in tasks that require fine synchronization. This invention optimizes the control algorithm, takes into account the interaction force and motion planning between the robotic arms, ensures that multiple robotic arms can achieve seamless collaboration when performing tasks, and maintain a high degree of consistency even in complex dynamic environments, thereby improving the collaborative operation capability of the overall system.

[0053] 3. Improve real-time performance and computing efficiency, and reduce dependence on precise models

[0054] Although model predictive control can handle constraints, it has a large amount of calculation and may not meet the requirements of real-time control, especially when dealing with large multi-manipulator systems. The present invention not only reduces the demand for computing resources and improves the real-time response capability of the system by designing efficient adaptive observers and control strategies, but also reduces the dependence on accurate model information. Even when there is a deviation between the model and the actual system, it can still maintain good control performance.

[0055] The detailed design plan is as follows:

[0056] 1. Initialization phase

[0057] Set the dynamic model of the robot arm, including parameters such as mass, inertia, and friction coefficient.

[0058]

[0059] Among them, M i is the moment of inertia, q i is the angular position, Relative velocity deflection angle, Angular acceleration, m i The mass of the connecting rod, g = 9.8 m / s 2 is a positive constant, l i The length of the robot arm, ζ i is the control torque, θ i Indicates input delay, i is the disturbance acting on the manipulator.

[0060] Define x i1 =q i , u i =ζ i (θ i ), a nonlinear system consisting of N+1 robotic arms, including N follower robotic arms and 1 leader robotic arm. The leader index is 0, and the follower indexes range from 1 to N. Then the system can be viewed as the following system:

[0061]

[0062] Among them, x i ∈R n ,u i ∈R p ,y i ∈R q , respectively, system state, control input, and output. State variables are not measurable, only output can be obtained by measurement, and u0(θ0(t))=0. θ i (t) is a smooth function and satisfies θ i (t) = t - τ i (t), τ i (t) is the non-uniform time-varying input delay. i (t):R + →R n Represents uncertain external disturbance. A∈R n×n , B∈R n×p , C∈R q×n is a constant matrix, (A,C) is observable, and (A,B) is controllable. i :R n →R n is an uncertain smooth nonlinear vector function.

[0063] definition in, is x i The estimated value of will be designed and implemented in the state observer, using radial basis function neural networks (RBFNNs) to analyze the uncertain nonlinear function Make an approximation We can get:

[0064]

[0065] in, is considered as a composite disturbance. and is a positive constant. The system model (2) can be rewritten as:

[0066]

[0067] 2. State Observation and Parameter Estimation

[0068] A consensus control strategy is constructed for each robot arm. The scheme consists of three observers based on adaptive RBFNNs, namely, the composite disturbance observer Local State Observer and distributed observers

[0069] (1) Design a composite disturbance observer to estimate the composite disturbance and define the following auxiliary variables:

[0070]

[0071] Among them, ρ i is a normal number.

[0072] According to formulas (4) and (5), we can get:

[0073]

[0074] Next, define the estimate of the composite perturbation:

[0075]

[0076] in, For composite disturbance The estimate, is an auxiliary variable The estimate, is the optimal weight matrix W i * Estimates.

[0077] Define the estimated error variable Combining formula (7), we can get:

[0078]

[0079] in,

[0080] right Taking the derivative and combining (6) and (7), we can obtain:

[0081]

[0082] in, is the weight matrix W i * The estimation error.

[0083] (2) Design a local state observer for each follower robot As shown below:

[0084]

[0085] in, is the state x i The estimate, is the output of the state observer, K∈R n×q is a gain matrix such that A+KC is a Hurwitz matrix.

[0086] The local state observation error is defined as We can get:

[0087]

[0088] in,

[0089] Positive design parameters γ, The following inequality group is satisfied:

[0090]

[0091] Weight matrix W * The adaptive law is:

[0092]

[0093] Among them, μ i is the design parameter.

[0094] and meet Where λ1 is The minimum eigenvalue of The minimum value of λ3 is The minimum value of .

[0095] (3) Design the following distributed observer based on output feedback for each follower:

[0096]

[0097] in, represents an estimate of the leader's state, represents the output of the distributed observer, Λ∈R q×n is the gain matrix, e i is the consistent tracking error, which is in the form of:

[0098]

[0099] Among them, x0 is the leader state, j represents the index of agent j; N i represents the neighboring agent of agent i; a ij represents the communication weight between agent i and agent j; b i represents the communication weight between agent i and the leader; y0 represents the leader output. A robotic arm is an agent.

[0100] The distributed observation error is defined as right Taking the derivative, we get:

[0101]

[0102] The leader's state can be estimated asymptotically by the distributed observer (15) if the following inequality is satisfied: in, is the upper boundary.

[0103]

[0104] in, is a positive definite matrix and γ is a positive constant. It can achieve semi-global eventually uniform boundedness, that is, λ z Yes (Q z -γ||P z ||)'s minimum eigenvalue.

[0105] 3. Controller design

[0106] By definition and Available Based on the above analysis, we have and is eventually uniformly bounded, that is, when t→∞, Therefore, it can be seen that if is eventually uniformly bounded. Through two types of observers (7) and (11), the nonlinear system (2) can achieve consistent tracking control.

[0107] definition The consistency tracking control problem of nonlinear system (2) is transformed into the stability analysis of the following nonlinear system:

[0108]

[0109] in,

[0110] The controller design is as follows:

[0111]

[0112] Among them, Γ i =-B T P i , consistency tracking control can be achieved. At the same time, the following conditions are met:

[0113]

[0114] Among them, P i is a positive definite matrix, Y i =P i -1 ,

[0115] In the control mode, the practical information is as follows:

[0116] 1. Parameter selection of adaptive observer: Select parameters θ i =1,γ i =0.04, β i =4, ι1=1.25, ι2=1.05. According to B and β i , select the matrix S i =[1.4869 -4.1721]. Let parameter K = [1,2] T , Λ=[1,2] T ,γ=0.4,

[0117] The initial state is x0(0) = [0.4, 0.2] T , x1(0)=[1.2,2.3] T , x2(0)=[-2.1,1.3] T , x3(0)=[1.2,-2.5] T , x4(0)=[2.1,1.3] T ,

[0118] A radial basis neural network with 9 neurons in the hidden layer is used. represents the basis function vector, where Activation function width μ i =1, center c i Uniformly distributed in the interval [-2,2]×[-2,2].

[0119] Choosing an appropriate initial estimate is crucial to the convergence speed and accuracy of the observer. Experimental data show that the initial estimate should be close to 50% to 80% of the true value, and the learning rate should be between 0.01 and 0.1 to balance the needs of fast response and stability.

[0120] 2. Optimization of control law design: The design of the control law should take into account the physical limitations of the robot, such as maximum torque and acceleration. i =[x i1 ,x i2 ] T ∈R 2 , f i (x i )=[0,-m i gl i sin(x i1 ) / 2M i ] T , d i (t)=[0,ò i / M i ] T , C=[1,0],B=[0,1 / M i ] T Design parameters i =0.2m,m i =0.3kg, M i =1kg.m,θ i =t-(0.01(t+1) / t+1), i =0.26sin(x i1 ), ò0=0, u0(θ0)=0. By introducing soft boundaries and penalty terms, the control signal can be prevented from exceeding the safety range while ensuring that the robot can track the target trajectory quickly and smoothly.

[0121] 3. Collaborative control strategy: In a multi-robot system, using the concepts of graph theory and network theory to define the connection relationship and information flow between the robots will help optimize the collaborative control strategy. Figure 2 In the directed communication topology shown, the application example consists of 5 robots, indexed 0-4.

[0122] Among them, we can get B = [1 0 0 1],

[0123] 4. Experimental verification: Build a simulation model of the multi-manipulator system on the MATLAB / Simulink platform, including tracking accuracy, response time, and computational load.

[0124] Through the above improvements, the present invention provides a more robust, accurate and efficient multi-arm tracking control method, which is particularly suitable for fields such as industrial automation, aerospace and medical surgery that require high precision and real-time response, and brings significant progress to the control technology of multi-arm systems.

[0125] The beneficial effects of the present invention are as follows:

[0126] 1. Significantly improve tracking accuracy: Through dynamic parameter estimation of the adaptive observer, the present invention can compensate in real time for tracking errors caused by parameter uncertainty (such as the inertia of the robot arm, changes in the friction coefficient) and external interference (such as wind force and vibration), so that the robot arm can follow the predetermined trajectory more accurately and improve the operation accuracy.

[0127] 2. Enhanced robustness and adaptability: The introduction of adaptive observers enhances the robustness of the control system, enabling it to maintain stable performance in the face of various unknown or changing environmental conditions. This is crucial for applications in industrial automation, aerospace, medical surgery, etc., because these scenarios often contain a large number of unpredictable factors.

[0128] 3. Optimize collaborative working capabilities: The present invention specially designs a collaborative control strategy that enables multiple robotic arms to maintain coordinated movements when performing complex tasks, reduces the risk of collision, and improves the overall operating efficiency and safety of the multi-robotic arm system.

[0129] 4. Improve real-time performance and computing efficiency: Compared with model predictive control, the method of the present invention reduces the burden of real-time calculation while ensuring the control effect, and improves the response speed of the system, which is especially important for scenarios that require rapid response.

[0130] 5. Economic benefits: The implementation of the present invention can significantly improve production efficiency, reduce scrap rate, and reduce maintenance costs. In the field of medical surgical robots, higher precision means less trauma and faster recovery, which brings better treatment experience to patients and also reduces medical costs.

[0131] An embodiment of the present invention provides a control system for a multi-manipulator system, including:

[0132] The system building module is used to build a nonlinear system in which a leader robot arm leads multiple follower robot arms to move through information interaction. The control design module is used to estimate the composite disturbance state of each robot arm using a composite disturbance observer for each robot arm in the nonlinear system; use a local state observer and a distributed observer for each robot arm to obtain the local state observation error and distributed observation error of each robot arm; determine the consistent bounded state of each robot arm based on the local state observation error and the distributed observation error. The control system building module is used to keep the nonlinear system running stably based on the composite disturbance state and the consistent bounded state, and can realize the consistent tracking control of the nonlinear system. The consistent tracking control is used to indicate that all robots can achieve consistent motion.

[0133] An embodiment of the present invention provides a computer device, comprising: a memory and a processor; the memory stores a computer program, and the processor implements the steps of a control method for a multi-robot system when executing the computer program.

[0134] An embodiment of the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of a control method for a multi-robotic arm system are implemented.

[0135] A specific embodiment is as follows:

[0136] 1. Dynamic parameter adaptation mechanism: By updating the dynamic parameter estimation of the robot arm in real time, the present invention can automatically adapt to changes in the working environment, reduce dependence on external calibration, and reduce maintenance costs.

[0137] 2. Collaborative control optimization algorithm: The collaborative control strategy of the present invention utilizes advanced optimization algorithms to intelligently adjust the relative position and movement between robotic arms, ensuring the efficiency and safety of the multi-robotic arm system when performing tasks, and opening up a new path for the automated execution of complex tasks.

[0138] 3. Economic benefit analysis: In the field of industrial manufacturing, higher precision and efficiency means increased output of the production line and reduced costs. In the medical field, the application of this invention can reduce operation time and complications, improve patient satisfaction, save costs for medical institutions, and improve the quality and level of medical services. These factors jointly promote the dual improvement of economic and social benefits of related industries.

[0139] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A control method for a multi-manipulator system, It is characterized in that The following steps are involved: Construct a nonlinear system in which a leader robot leads multiple follower robots to move through information interaction; In the nonlinear system, a composite disturbance observer is used for each manipulator to estimate the composite disturbance state of each manipulator; a local state observer and a distributed observer are used for each manipulator to obtain the local state observation error and the distributed observation error of each manipulator; the consistently bounded state of each manipulator is determined based on the local state observation error and the distributed observation error; The nonlinear system is kept running stably according to the composite disturbance state and the consistent bounded state, and the consistent tracking control of the nonlinear system can be realized. The consistent tracking control is used to indicate that all the robotic arms can realize consistent motion.

2. A control method for a multi-manipulator system as claimed in claim 1, It is characterized in that The specific steps of obtaining the local state observation error and the distributed observation error of each robotic arm include: Use a local state observer for each robot The local state observation error is defined as Where N represents the number of follower robotic arms, i represents the i-th follower robotic arm, is the state x i The estimate, is the output of the state observer, u i ∈R p is the control input, K∈R n×q is the gain matrix, making A+KC a Hurwitz matrix; θ i (t) represents the input delay and satisfies θ i (t) = t - τ i (t), τ i (t) is the non-uniform time-varying input delay; represents uncertain external disturbance; and are the weight matrix and activation function of the radial basis neural network, A∈R n ×n , B∈R n×p , C∈R q×n is a constant matrix; Use a distributed observer for each robot in, represents an estimate of the leader's state, represents the output of the distributed observer, Λ∈R q×n is the gain matrix, e i is the consistent tracking error, and the formula is: The distributed observation error is defined as Among them, x 0 is the leader state, j represents the index of agent j; N i represents the neighboring agent of agent i; a ij represents the communication weight between agent i and agent j; b i represents the communication weight between agent i and the leader; y 0 Represents the leader output.

3. A control method for a multi-manipulator system as claimed in claim 1, It is characterized in that The steps of determining the consistent bounded state of each robot arm according to the local state observation error and the distributed observation error include: Get local state observation error and distributed observation errors According to the two errors When time t tends to infinity, we get and Determine the consistent and bounded state of each robot.

4. A control method for a multi-manipulator system as claimed in claim 1, It is characterized in that The specific steps of realizing the consistency tracking control of the nonlinear system include: The consistency tracking control problem of nonlinear system is transformed into the stability analysis of nonlinear system through composite disturbance observer, local state observer and distributed observer, and the stability model of nonlinear system is obtained. The stability model of the nonlinear system is as follows: in, Uncertain external disturbances; Local state observation error; is the estimation error, and They are the weight matrix and activation function of the radial basis neural network, Represents the estimate of the leader state; A∈R n×n , B∈R n×p , C∈R q×n is a constant matrix, K∈R n×q ,Λ∈R q×n is the gain matrix; e i is the consistent tracking error; u i ∈R p is the control input; θ i Indicates input delay; Error information, Follower state state estimation, represents the leader state estimate; Building a controller according to the stability model of the nonlinear system, wherein the controller can realize the consistency tracking control of the nonlinear system; The controller formula is: Among them, u i ∈R p is the control input; Γ i =-B T P i , P i is a positive definite matrix; e i is the consistent tracking error; A∈R n×n is a constant matrix; The input delay boundary is a known constant; Error information, Follower state estimation, Leader state estimation.

5. A control system for a multi-manipulator system, It is characterized in that include: System building module, used to build a nonlinear system in which a leader robot leads multiple follower robots to move through information interaction; A control design module is used to estimate the composite disturbance state of each manipulator using a composite disturbance observer for each manipulator in a nonlinear system; obtain the local state observation error and the distributed observation error of each manipulator using a local state observer and a distributed observation error for each manipulator respectively; and determine the consistent bounded state of each manipulator according to the local state observation error and the distributed observation error; The control system building module is used to keep the nonlinear system running stably according to the composite disturbance state and the consistent bounded state, and can realize the consistent tracking control of the nonlinear system, wherein the consistent tracking control is used to indicate that all the robotic arms can realize consistent motion.

6. A computer device, include: Memory and processor; The memory stores a computer program, wherein the processor implements a control method for a multi-robotic arm system according to any one of claims 1 to 4 when executing the computer program.

7. A computer-readable storage medium storing a computer program, It is characterized in that When the computer program is executed by a processor, a control method for a multi-manipulator system according to any one of claims 1 to 4 is implemented.

Images