Preset time event triggering control method for flexible joint manipulator under error constraints

Through the flexible joint robot arm adaptive preset time event trigger control method with full state error constraint, the use of obstacle Liyapunov function and event triggering strategy, the shortcomings of traditional control methods in high-precision and fast response scenarios are solved, and the system is realized The efficient utilization of resources and real-time performance ensures error convergence and trajectory tracking effects.

CN118357919BActive Publication Date: 2024-12-06NINGXIA WUCHANGJIAN MEDICAL EQUIPMENT CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410608789.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-05-16
Publication Date
2024-12-06
Estimated Expiration
2044-05-16

AI Technical Summary

Technical Problem

Traditional limited time control and fixed time control methods are difficult to meet the control performance requirements of flexible joint robotic arm systems in high-precision and fast response scenarios, and periodic sampling leads to resource waste and real-time delays.

Method used

Adaptive preset time event trigger control method of flexible joint robot arm with full state error constraints, using obstacle Liyapunov function and dynamic surface technology, optimize resource utilization through event triggering strategies, build preset filters to avoid differential explosions, and according to Lyapunov's stability theory achieves preset time stability.

Benefits of technology

The convergence of full state error and tracking error to the user-specified accuracy range is achieved, which improves resource utilization efficiency and system real-time and response performance, avoids excessive resource consumption, and ensures the system's preset time stability and tracking effect. .

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118357919B_ABST
    Figure CN118357919B_ABST
Patent Text Reader

Abstract

The present invention proposes a preset time event-triggered control method for a flexible joint manipulator under error constraints. The study uses a radial basis function neural network (RBFNN) to process unknown nonlinear dynamic functions. Based on the event-triggered control strategy, a communication-restricted control scheme is established to reduce the execution time of the controller and save communication resources. Combined with the dynamic surface (DSC) technology, the differential explosion problem is avoided by using the constructed preset time filter. By using the proposed preset time stability criterion and the universal barrier Lyapunov function, it can be proved that the controlled system achieves semi-global practical preset time stability (PPTS), and all signals are within the preset time interval. Finally, the effectiveness of the method is verified by simulation experiments.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of robot arm control, and in particular relates to a preset time event triggering control method of a flexible joint robot arm under error constraints. Background Art

[0002] In recent years, flexible robotic arms have become a popular choice in various fields due to their lightweight design, low energy consumption, and safety in human-machine interaction. These robotic arms can flexibly adapt to different working environments and can reduce potential safety risks when working with humans. Their design enables them to operate in narrow spaces and perform various tasks, such as medical surgery, warehouse management, and disaster relief. With the continuous development of science and technology, the application prospects of flexible robotic arms will be broader, bringing more convenience and benefits to mankind.

[0003] However, the complexity of these system dynamics and the strong coupling and incompleteness of most underactuated systems make the design of their controllers more challenging. A large number of studies have focused on the tracking control of flexible joint manipulator systems, including sliding mode control, variable structure control, backstepping control, etc. However, the existing technologies still have the following problems:

[0004] First, traditional finite-time control and fixed-time control methods may not guarantee good control performance in some cases. For example, in scenarios where high-precision control or high control speed are required, these methods may not meet the requirements. These two methods are sensitive to changes in system parameters, especially for nonlinear systems or systems with uncertainties. Changes in parameters may cause control performance to deteriorate or system instability.

[0005] Second, traditional periodic sampling and control may lead to waste of resources, especially when the system state changes slowly or is stable. In this case, periodic sampling may lead to a large amount of repeated information transmission and processing, increasing the communication and computing costs of the system. In addition, periodic sampling may introduce inherent delays, thus affecting the real-time and response performance of the system. Especially in real-time control systems that require fast response, fixed-frequency periodic sampling may not meet the real-time requirements. Summary of the invention

[0006] In order to solve the deficiencies in the prior art, the present invention designs an adaptive preset time event trigger control method for a flexible joint robot arm with full-state error constraints. The technical solution of the present invention is as follows: By using an obstacle Lyapunov function and dynamic surface technology, the full-state error and tracking error are converged to the control accuracy range specified by the user, and the constructed preset filter can avoid the differential explosion problem. Through the event triggering strategy, the utilization of resources is optimized, and sampling and control actions are selectively triggered according to the changes in the system state, thereby improving the utilization efficiency of resources. Finally, according to the Lyapunov stability theory and the definition of preset time, it is proved that the system achieves actual preset time stability.

[0007] Step 1, establish a flexible joint robot model and rewrite it into a state space expression;

[0008] Step 2: Introduce the obstacle Lyapunov function to solve the full-state error constraint problem. Convert the tracking error into a conversion error through coordinate transformation to ensure that the virtual tracking error z i Converge to the user-defined constraint function X H,i , X L,i within the scope;

[0009] Construct Lyapunov functions, design virtual control signals and adaptive laws;

[0010] Step 3, construct a controller and event trigger strategy; input the virtual control signal into the controller, combine it with the event trigger strategy, and output the control signal u(t) to control the flexible joint robot arm.

[0011] Furthermore, the model equation of the flexible joint robot arm established in step 1 is described as follows;

[0012]

[0013] in, Represent the chain link position, velocity and acceleration vector respectively, M(q)∈R n×n is the inertia matrix, which is a symmetric positive definite matrix; is the Coriolis force and centripetal force matrix, G(q)∈R n is the gravity vector, represents friction; Represent the rotor angular position, velocity and acceleration vector respectively; the positive definite matrix K∈R n×n , J∈R n×n and B∈R n×n denote the joint flexibility, actuator inertia and natural damping term respectively;

[0014] Let x1=q=[x 1,1 ,x 1,2 ,...,x 1,n] T , x3=q m =[x 3,1 ,x 3,2 ,...,x 3,n ] T ,

[0015] The robot model equation (1) is rewritten as a state space expression:

[0016]

[0017] in,

[0018]

[0019] In order to approximate f2 and f4 in the state space expression, a radial basis neural network is established to approximate the unknown nonlinear function:

[0020]

[0021] in, is the basis function vector, j>1 represents the number of nodes; in addition, the upper limit of the approximation error ε is

[0022] The ideal weight vector is:

[0023]

[0024] Where W=[W1,...,W j ] T ;

[0025] The construction of radial basis function neural network is as follows:

[0026]

[0027] The center is Width x i =[x i,1 ,...,x i,v ];

[0028] In order to reduce the amount of calculation, Young's inequality is used for scaling:

[0029]

[0030] Let θ = || W *T || 2 And k θ Is a normal number.

[0031] Furthermore, in step 2, the barrier Lyapunov function is introduced to solve the full-state error constraint problem. The tracking error is converted into a transformation error through coordinate transformation to ensure that the virtual tracking error z i Converge to the user-defined constraint function X H,i , X L,i The specific steps are as follows:

[0032]

[0033] Among them, x i is the system state; y=x1 is the system output state; y r is the specified reference signal; z1 is the system tracking error; z i is the virtual tracking error; s i is the filtering error; is the filtered output signal; α i is a virtual control signal;

[0034] Then use the barrier Lyapunov function to convert it into the conversion error ψ i The tracking is as follows:

[0035]

[0036] -X L,i (0) <z i (0) <X H,i (0) (33)

[0037] Among them, ψ i is the conversion error; X H,i and X L,i It is a user-defined time-varying constraint function that is required to satisfy -X L,i (0) <z i (0) <X H,i (0) condition.

[0038] Furthermore, combined with the barrier Lyapunov function, the coordinate transformation is:

[0039]

[0040] in,

[0041]

[0042] Furthermore, the Lyapunov function constructed in step 2 is:

[0043]

[0044]

[0045]

[0046]

[0047] Furthermore, the virtual control signal expression in step 2 is:

[0048]

[0049]

[0050]

[0051]

[0052] The constructed adaptive law is:

[0053]

[0054]

[0055]

[0056] Furthermore, a preset time filter is constructed in step 2 to avoid the problem of differential explosion;

[0057] It is represented by the following differential equation:

[0058]

[0059]

[0060] where and h = h1 / h2, h1 and h2 are positive even and odd integers satisfying h1 < h2, T d is a positive design parameter representing the preset time, υ i represents a positive constant; where, ω i ≥ 1 and υ i > 0 are design parameters.

[0061] Furthermore, the controller in step 3 is:

[0062]

[0063] where and l(t) is a time-varying parameter satisfying The event-triggering strategy refers to:

[0064]

[0065]

[0066] Wherein, v(t) represents the continuous control signal, P(t) = v(t) - u(t); <d<1,Z1> 0, Z2>0 and Ξ>0 are design parameters.

[0067] Compared with the prior art, the present invention has the following beneficial effects:

[0068] First, this paper proposes an adaptive preset time control method for a flexible joint robot arm system. Unlike traditional finite time control and fixed time control, the preset time given by the present invention is not only independent of the initial value, but also does not require the adjustment of other parameters. Compared with finite time control, preset time control can plan and predict the execution time of tasks in advance, which is helpful for the reasonable planning and resource allocation of the system. Compared with fixed time control, preset time control can control the execution time of tasks more accurately and improve the accuracy and reliability of the system.

[0069] Second, based on the preset time stability theory and barrier Lyapunov function, a nonlinear system adaptive control method with full-state error constraint is proposed. Through the event trigger mechanism given in this paper, the controller can avoid excessive consumption of communication and computing resources. In addition, this paper also uses the barrier Lyapunov function to give a general full-state error constraint solution, constraining the full-state tracking error within the user-specified constraint function. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 It is a schematic diagram of the network system of the full coefficient identification method of the flexible joint mechanical model.

[0071] Figure 2 This is the trajectory tracking effect diagram of the flexible joint robotic arm model.

[0072] Figure 3 It is the effect diagram of the state tracking error trajectory z1 of the flexible joint robot arm model system.

[0073] Figure 4 It is the effect diagram of the state tracking error trajectory z2 of the flexible joint robot arm model system.

[0074] Figure 5 It is the effect diagram of the state tracking error trajectory z3 of the flexible joint robot arm model system.

[0075] Figure 6 It is the effect diagram of the state tracking error trajectory z4 of the flexible joint robot arm model system.

[0076] Figure 7 It is the effect diagram of the filtering errors s1, s2 and s3 of the flexible joint robot model.

[0077] Figure 8 This is the effect diagram of the controllers u and v of the flexible joint robot model.

[0078] Fig. 9 This is the effect diagram of the event trigger interval of the flexible joint robot arm model. DETAILED DESCRIPTION

[0079] The preset time event triggering control method of the flexible joint mechanical arm under error constraint of the present invention comprises the following steps:

[0080] A. Establish a flexible joint robotic arm model

[0081] The robot arm model is described by the following equation:

[0082]

[0083] in, Represent the chain link position, velocity and acceleration vector respectively, M(q)∈R n×n is the inertia matrix, which is a symmetric positive definite matrix. is the Coriolis force and centripetal force matrix, G(q)∈R n is the gravity vector, Represents friction. Represent the rotor angular position, velocity and acceleration vector respectively. The positive definite matrix K∈R n×n , J∈R n×n and B∈R n×n denote the joint flexibility, actuator inertia and natural damping terms respectively.

[0084] Let x1=q=[x 1,1 ,x 1,2 ,...,x 1,n ] T , x3=q m =[x 3,1 ,x 3,2 ,...,x 3,n ] T ,

[0085] The robot model equation (1) is rewritten as a state space expression:

[0086]

[0087] in,

[0088]

[0089] B. Define coordinate transformation

[0090]

[0091] Among them, x i is the system state; y=x1 is the system output state; y r is the specified reference signal; z1 is the system tracking error; z i is the virtual tracking error; s i is the filtering error; is the filtered output signal; α i is a virtual control signal that is used as the filter output signal.

[0092] C. Introducing radial basis neural network

[0093] Establish a radial basis neural network to approximate unknown nonlinear functions:

[0094]

[0095] in, is the basis function vector, j>1 represents the number of nodes. In addition, the upper limit of the approximation error ε is

[0096] The ideal weight vector is:

[0097]

[0098] Where W=[W1,...,W j ] T .

[0099] The construction of radial basis function neural network is as follows:

[0100]

[0101] The center is Width x i =[x i,1 ,...,x i,v ].

[0102] In order to reduce the amount of calculation, Young's inequality is used for scaling:

[0103]

[0104] Let θ = || W *T || 2 And k θ Is a normal number.

[0105] D. Non-singular preset time filter equation

[0106] The filter equation is expressed by the following differential equation:

[0107]

[0108]

[0109] Among them, and h = h1 / h2, where h1 and h2 are positive even and odd integers satisfying h1 < h2, and T d is a positive design parameter representing a preset time, and υ i represents a positive constant. Among them, ω i ≥ 1 and υ i > 0 are design parameters. In subsequent control design, the designed preset time filter will be used to process the virtual control signal.

[0110] E, barrier Lyapunov function

[0111] Introduce the following barrier Lyapunov function equation:

[0112]

[0113] Among them, ψ i is the conversion error; X H,i and X L,i are user-defined time-varying constraint functions, which are required to satisfy the condition of -X L,i (0) < z i (0) < X H,i (0).

[0114] Combine the barrier Lyapunov function and the coordinate transformation:

[0115]

[0116] Among them,

[0117]

[0118] F, construct Lyapunov function, virtual control signal and adaptive law

[0119] The constructed Lyapunov function is:

[0120]

[0121]

[0122]

[0123]

[0124] The constructed virtual control signal is:

[0125]

[0126]

[0127]

[0128]

[0129] The adaptive law constructed is used to improve the intelligence and adaptability of the system. The adaptive law is:

[0130]

[0131]

[0132]

[0133] G. Event-triggered strategies and controllers

[0134] When network communication resources and bandwidth are limited, the following event trigger strategies are available:

[0135]

[0136]

[0137] Wherein, v(t) represents the continuous control signal, P(t) = v(t) - u(t); <d<1,Z1> 0, Z2>0 and Ξ>0 are design parameters.

[0138] The constructed controller is:

[0139]

[0140] in, And l(t) satisfies time-varying parameters.

[0141] The present invention is further described below in conjunction with the accompanying drawings. The network system diagram of the adaptive preset time event triggering control method of the flexible joint robot model with full state error constraints of the present invention is shown in FIG. Figure 1 The simulation results are shown in Figure 2-6 shown. Figure 2 The trajectory tracking effect of the flexible joint robotic arm model is shown. Figure 3 The effect diagram of the state tracking error trajectory z1 of the flexible joint robot arm model system is shown. Figure 4 The effect diagram of the state tracking error trajectory z2 of the flexible joint robot arm model system is shown. Figure 5 The effect diagram of the state tracking error trajectory z3 of the flexible joint robot arm model system is shown. Figure 6 The effect diagram of the state tracking error trajectory z4 of the flexible joint robot arm model system is shown. Figure 7 The effect diagram of the filtering errors s1, s2 and s3 of the flexible joint robot model is shown. Figure 8 The effect diagram of the controllers u and v of the flexible joint robot model is shown. Fig. 9 The effect diagram of the event trigger interval of the flexible joint robot model is shown. It can be seen from the above simulation result diagram that under this strategy, the system can achieve actual preset time stability, good trajectory tracking effect, and the full state error and tracking error converge to the control accuracy range specified by the user.

[0142] The present invention is not limited to this embodiment, and any equivalent concepts or changes within the technical scope disclosed by the present invention are included in the protection scope of the present invention.

Claims

1. A preset time event triggering control method for a flexible joint manipulator under error constraints, characterized in that: The steps include: Step 1, establish a flexible joint robot model and rewrite it into a state space expression; Step 2, introduce the obstacle Lyapunov function and convert the tracking error into the conversion error through coordinate transformation; Construct Lyapunov functions, design virtual control signals and adaptive laws; A preset time filter is constructed, which is represented by the following differential equation: in, and , and is satisfied < positive even integers and odd integers, T d is a positive design parameter representing the preset time, represents a positive constant; where ω i ≥1 and >0 is the design parameter, s i is the filtering error; in, is the filtered output signal; α i is a virtual control signal; Step 3, construct a controller and an event trigger strategy; input the virtual control signal into the controller, combine it with the event trigger strategy, and output the control signal to control the flexible joint robot arm.

2. The preset time event triggering control method of a flexible joint manipulator under error constraints according to claim 1 is characterized in that The flexible joint manipulator model equation described in step 1 is described as follows; in, Represent the chain link position, velocity and acceleration vector respectively, M(q)∈R n×n is the inertia matrix, which is a symmetric positive definite matrix; is the Coriolis force and centripetal force matrix, G(q)∈R n is the gravity vector, represents friction; Represent the rotor angular position, velocity and acceleration vector respectively; the positive definite matrix K∈R n×n , J∈R n ×n and B∈R n×n denote the joint flexibility, actuator inertia and natural damping term respectively; Let x1 = q = [x 1,1 , x 1,2 ,..., x 1,n T , x3 = q m = [x 3,1 , x 3,2 ,..., x 3,n T , ​​ The robot model equation (1) is rewritten as a state space expression: in, Among them, y=x1 represents the system output state.

3. The preset time event triggering control method of a flexible joint manipulator under error constraints according to claim 2 is characterized in that: In order to approximate f2 and f4, a radial basis neural network is established to approximate the unknown nonlinear function: in, is the basis function vector, j>1 represents the number of nodes; in addition, the upper limit of the approximation error ε is The ideal weight vector is: Where W=[W1,...,W j ] T ; The construction of radial basis function neural network is as follows: The center is Width is ; In order to reduce the amount of calculation, Young's inequality is used for scaling: Let θ = || W *T || 2 And k θ Is a normal number.

4. The preset time event triggering control method of a flexible joint manipulator under error constraints according to claim 1 is characterized in that: In step 2, the barrier Lyapunov function is introduced to transform the tracking of the system state into the conversion error ψ through coordinate transformation. i The specific steps for tracking are as follows: Among them, x i is the system state; y=x1 is the system output state; y r is the specified reference signal; z1 is the system tracking error; z i is the virtual tracking error; s i is the filtering error; is the filtered output signal; α i is a virtual control signal; Reuse barrier Lyapunov function V b,i , which is converted to the conversion error ψ i The tracking is as follows: -X L,i (0)<z i (0)<X H,i (0) (10) Among them, ψ i is the conversion error; X H,i and X L,i It is a user-defined time-varying constraint function that is required to satisfy -X L,i (0) <z i (0) <X H,i (0) condition.

5. The preset time event triggering control method of a flexible joint manipulator under error constraints according to claim 4 is characterized in that: Combined with the barrier Lyapunov function, the coordinate transformation is: in, 6. The preset time event triggering control method of a flexible joint manipulator under error constraints according to claim 5 is characterized in that: The Lyapunov function constructed in step 2 is:

7. The preset time event triggering control method of a flexible joint manipulator under error constraints according to claim 6 is characterized in that: The virtual control signal expression in step 2 is: The constructed adaptive law is:

8. The preset time event triggering control method of a flexible joint manipulator under error constraints according to claim 1 is characterized in that: The controller described in step 3 is: in, And l(t) satisfies time-varying parameters.

9. The preset time event triggering control method of a flexible joint manipulator under error constraints according to claim 8 is characterized in that: The event triggering strategy described in step 3 refers to: Wherein, v(t) represents the continuous control signal, P(t)=v(t)-u(t); 0< <1,Z1>0,Z2>0 and >0 is a design parameter.