Self-adaptive inversion sliding mode control method for mechanical arm
By adopting an adaptive inversion sliding mode control method, the problem of accurate trajectory tracking of the robotic arm under external disturbances and frictional torques was solved, achieving high-precision and robust robotic arm control, reducing chattering, and ensuring system stability.
Patent Information
- Application Number
- CN202411002452.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-24
- Publication Date
- 2024-11-08
AI Technical Summary
In existing technologies, robotic arms struggle to achieve precise trajectory tracking when faced with external disturbances and joint friction, and sliding mode control systems are prone to chattering, affecting the accuracy and real-time performance of the robotic arm's movements.
An adaptive inversion sliding mode control method is adopted. By establishing a dynamic model of the robotic arm and combining it with the LuGre model to describe the friction torque, a disturbance observer is designed for observation and compensation. In addition, by combining an adaptive strategy and an inversion sliding mode controller, chattering is suppressed and the robustness of the system is improved.
It enables accurate observation and online estimation of unknown external disturbances, improves the anti-interference capability and trajectory tracking accuracy of the robotic arm system, reduces chattering, and ensures the global asymptotic stability of the system.
Smart Images

Figure CN118915459A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of robot arm control, and in particular to a robot arm adaptive inversion sliding mode control method. Background Art
[0002] The robot arm is a kind of dynamic system with strong nonlinearity, strong coupling, and many model parameters, which has strong uncertainty. With the rapid development of science and technology, the robot arm is more and more widely used in various industrial production. In many practical applications, accurate trajectory tracking of the robot arm is very important. The existence of objective factors such as external disturbances and joint friction will affect the accuracy and real-time performance of the robot arm joint movement, thereby increasing the difficulty of tracking and controlling the trajectory of the end of the robot arm.
[0003] In recent years, domestic and foreign scholars have made many explorations on the trajectory tracking control of manipulators, and summarized a large number of theoretical results in the optimization of control algorithms, control parameters and disturbance compensation. Due to the strong robustness of sliding mode control, the research on this control method has gradually become a hot direction in recent years. However, when the control system state converges to the vicinity of the sliding surface, it is difficult to ensure that the system will run completely along the set sliding surface, which will cause the control system to vibrate. In order to enhance the robustness of the controller, a control method based on disturbance observer is proposed. This method can help the manipulator system estimate unknown disturbances, uncertain parameters and unknown states, thereby improving the trajectory tracking control accuracy of the manipulator system. The sliding mode control scheme based on disturbance observer can better solve the chattering phenomenon in sliding mode control, thereby improving the anti-interference ability of the system. However, the disturbance observer generally cannot accurately observe the impact of joint friction on the real-time and stability of the entire manipulator system. Joint friction has a great influence on the motion performance of the manipulator. With the increase in the requirements for control accuracy, it is very important to solve the impact of joint friction on the trajectory tracking control accuracy of the manipulator. Summary of the invention
[0004] The purpose of the present invention is to provide a method for adaptive backstepping sliding mode control of a robotic arm, which can accurately observe and online estimate external unknown disturbances and improve the anti-interference ability of the robotic arm system.
[0005] In order to achieve the above object, the technical solution adopted by the present invention is: a method for adaptive backstepping sliding mode control of a manipulator, comprising the following steps: Step 1: Establish an n-dimensional robotic arm dynamics model based on Lagrangian dynamics mechanism; Step 2: Considering that the disturbance observer cannot accurately describe the joint friction torque, the dynamic model of the manipulator is modified and the mathematical model of the dynamics of the manipulator is converted into a state space equation form; Step 3, using the LuGre model to describe the friction torque that the disturbance observer cannot accurately observe, and performing offline identification of the friction parameters; Step 4: Design a disturbance observer to observe and compensate for uncertain parameters and unknown disturbances; Step 5: Design a sliding mode controller by combining the adaptive strategy and the backstepping sliding mode control; Step 6: Use the Lyapunov stability criterion to analyze the stability of the robotic arm system.
[0006] Preferably, the mechanical arm dynamics model in step 1 is as follows: Where: M(q)∈R n×n is the positive definite inertia matrix of the robotic arm system; is the centripetal force and Coriolis force matrix of the robotic arm system; G(q)∈R n is the gravity vector acting on the robot joint; u∈R n Control input torque for the robot system; is the total unknown disturbance of the robotic system, where τ d is the unknown time-varying disturbance, ΔM, ΔC, ΔG are the modeling errors of the robot system; q, are the actual angle, angular velocity and angular acceleration of the robot.
[0007] Preferably, the specific steps of modifying the mechanical arm dynamics model and converting the mechanical arm dynamics mathematical model into a state space equation form in step 2 are as follows: For the joint friction torque F, F will switch between positive and negative when the robot joint turns. The first-order derivative of the joint friction torque at this time is Considering the joint friction torque as part of the total disturbance f, at this stage The joint friction torque cannot be simply regarded as part of the total disturbance f; in order to accurately feedforward compensate the robot system, the robot dynamics model can be written as follows: Where: f1 is the total disturbance excluding the joint friction torque; F is the joint friction torque; Let x1=q, Then the state space equation of the robotic arm system can be expressed as:
[0008] Preferably, the LuGre friction torque model in step 3 is as follows: Where: v represents the relative speed of the two objects; v sis the Stribeck speed effect; T c 、T s are Coulomb friction and static friction respectively; σ0 is the stiffness coefficient of the bristle; σ1 is the microscopic damping coefficient of the bristle; σ2 is the viscous damping coefficient; σ0 and σ1 are dynamic friction parameters; g(v) is a velocity-related function; z is the average deformation of the bristle.
[0009] Preferably, in step 3, a constant speed tracking experiment is performed to collect raw friction data, and the friction parameters are identified using identification technology. When the joint of the manipulator moves at a constant speed, the friction torque can be described as follows: f v =g(v)sgnv+σ2v.
[0010] Preferably, in step 4, it is assumed that the total unknown disturbance f satisfies continuous differentiability and boundedness and the bounds are unknown, that is, Based on the dynamic model in step 1, a disturbance observer is designed to observe the system disturbance; the disturbance observer is as follows: in: is the nonlinear function to be designed; is the friction torque of the joint after identification; is the gain of the disturbance observer, is the estimated value of f; After adopting the disturbance observer, the robotic arm system in step 2 becomes in: is the observation error of the disturbance observer; u BS is the sum of equivalent control and switching control.
[0011] Preferably, after introducing the disturbance observer in step 5, the state space equation in step 2 is written as: Combined with the relevant theories of sliding mode control, the adaptive backstepping sliding mode controller for the manipulator is designed in two steps.
[0012] Preferably, in step 5, the mechanical arm adaptive backstepping sliding mode controller is designed in two steps in combination with the relevant theories of sliding mode control. The specific steps are as follows: Step 51: Set q d Given the position command for the robot arm, z1 is defined as the control output error of the system, that is, z1 = qq d Define the Lyapunov function: where c1∈R n×nis a positive definite symmetric constant matrix; definition: Taking the derivative of the Lyapunov function, we get: If z2=0, then is the quadratic function of the system's control output error z1, when The next step of design is required; Step 52: Define the sliding surface function: s = k1z1 + z2 Where: k1∈R n×n is a symmetric positive definite constant matrix; Define the Lyapunov function: Taking the derivative of the Lyapunov function, we get: Based on the previous backstepping sliding mode controller design, an adaptive algorithm is used to adjust the unobservable part of the external disturbance signal. Make an estimate; define For the unobservable part The estimated value of is defined as: Define the Lyapunov function: Among them: γ is a positive constant; after taking the derivative of the Lyapunov function, we can get: The control law designed by the above formula is as follows: Where: u eq is equivalent control; u vs is the switching control; h∈R n×n is a symmetric, positive constant matrix; β is a positive constant; sgn is a sign function; The adaptive law can be taken as:
[0013] Preferably, in step 6, for the entire closed-loop system, the Lyapunov function can be written as follows: Taking the derivative of the Lyapunov function, we get: The above formula can be written as follows: in:
[0014] Preferably, in order to The matrix P needs to be a positive definite matrix to ensure that the robotic arm system meets the Lyapunov stability criterion.
[0015] The beneficial effects of the present invention are:
[0016] This scheme designs a disturbance observer to observe and compensate for the disturbances such as uncertain parameters and unknown disturbances. It can accurately observe and estimate the external unknown disturbances online, and improve the anti-interference ability of the robot system. Compared with the traditional method of using only disturbance observers, this paper takes into account the influence of nonlinear factors such as friction, and uses identification technology to identify friction, which improves the control accuracy of given trajectory tracking. In order to suppress the "jittering" of the system, the sliding mode controller is designed by combining adaptive and inverse sliding mode control strategies. On the basis of the design of the inverse sliding mode controller, the adaptive control theory is used to compensate for the external disturbance signal of the robot, which improves the robustness and convergence speed of the system. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0018] Figure 1 is a control block diagram of the inverse sliding mode controller in the present invention;
[0019] Figure 2 Schematic diagram of a two-degree-of-freedom rigid robotic arm model in an embodiment of the present invention;
[0020] Figure 3 Schematic diagram of robot arm joint position tracking under the experimental group method in an embodiment of the present invention;
[0021] Figure 4 Schematic diagram of robot arm joint position tracking under the control group method in an embodiment of the present invention;
[0022] Figure 5 Schematic diagram of the robot arm joint position tracking error under the experimental group and control group methods in the embodiment of the present invention;
[0023] Figure 6 Schematic diagram of robot arm joint speed tracking under the experimental group method in an embodiment of the present invention;
[0024] Figure 7 Schematic diagram of robot arm joint speed tracking under the control group method in an embodiment of the present invention;
[0025] Figure 8Schematic diagram of robot arm joint velocity tracking error under the experimental group and control group methods in an embodiment of the present invention;
[0026] Fig. 9 Schematic diagram of the comparison of mechanical arm joint torques between the experimental group and the control group in an embodiment of the present invention. DETAILED DESCRIPTION
[0027] The present invention is further described in detail below in conjunction with embodiments and with reference to the accompanying drawings.
[0028] The present invention discloses a method for adaptive backstepping sliding mode control of a mechanical arm. Figure 1-9 As shown, the embodiment includes the following steps: Step 1: Establish an n-dimensional robotic arm dynamics model based on Lagrangian dynamics mechanism; Step 2: Considering that the disturbance observer cannot accurately describe the joint friction torque, the dynamic model of the manipulator is modified and the mathematical model of the dynamics of the manipulator is converted into a state space equation form; Step 3, using the LuGre model to describe the friction torque that the disturbance observer cannot accurately observe, and performing offline identification of the friction parameters; Step 4: Design a disturbance observer to observe and compensate for uncertain parameters and unknown disturbances; Step 5: Design a sliding mode controller by combining the adaptive strategy and the backstepping sliding mode control; Step 6: Use the Lyapunov stability criterion to analyze the stability of the robotic arm system.
[0029] The dynamic model of the robotic arm in step 1 is as follows: Where: M(q)∈R n×n is the positive definite inertia matrix of the robotic arm system; is the centripetal force and Coriolis force matrix of the robotic arm system; G(q)∈R n is the gravity vector acting on the robot joint; u∈R n Control input torque for the robot system; is the total unknown disturbance of the robotic system, where τ d is the unknown time-varying disturbance, ΔM, ΔC, ΔG are the modeling errors of the robot system; q, are the actual angle, angular velocity and angular acceleration of the robot.
[0030] The specific steps of modifying the mechanical arm dynamics model and converting the mechanical arm dynamics mathematical model into the state space equation form in step 2 are as follows: For the joint friction torque F, F will switch between positive and negative when the robot joint turns. The first-order derivative of the joint friction torque at this time is Considering the joint friction torque as part of the total disturbance f, at this stage The joint friction torque cannot be simply regarded as part of the total disturbance f; in order to accurately feedforward compensate the robot system, the robot dynamics model can be written as follows: Where: f1 is the total disturbance excluding the joint friction torque; F is the joint friction torque; Let x1 = q, Then the state space equation of the robotic arm system can be expressed as:
[0031] The LuGre friction torque model in step 3 is as follows: Where: v represents the relative speed of the two objects; v s is the Stribeck speed effect; T c , T s are Coulomb friction and static friction respectively; σ0 is the stiffness coefficient of the bristle; σ1 is the microscopic damping coefficient of the bristle; σ2 is the viscous damping coefficient; σ0 and σ1 are dynamic friction parameters; g(v) is a velocity-related function; z is the average deformation of the bristle.
[0032] In step 3, a constant speed tracking experiment is performed to collect the original friction data, and the friction parameters are identified using identification technology. When the robot arm joint moves at a constant speed, the friction torque can be described as follows: f v =g(v)sgnv+σ2v.
[0033] In step 4, it is assumed that the total unknown disturbance f satisfies continuous differentiability and is bounded and the bound is unknown, that is, Based on the dynamic model in step 1, a disturbance observer is designed to observe the system disturbance; the disturbance observer is as follows: in: is the nonlinear function to be designed; is the friction torque of the joint after identification; is the gain of the disturbance observer, is the estimated value of f; After adopting the disturbance observer, the robotic arm system in step 2 becomes in: is the observation error of the disturbance observer; u BS is the sum of equivalent control and switching control.
[0034] After the disturbance observer is introduced in step 5, the state space equation in step 2 is written as: Combined with the relevant theories of sliding mode control, the adaptive backstepping sliding mode controller for the manipulator is designed in two steps.
[0035] In step 5, the related theories of sliding mode control are combined to design the adaptive backstepping sliding mode controller of the manipulator in two steps. The specific steps are as follows: Step 51: Set q d Given the position command for the robot arm, z1 is defined as the control output error of the system, that is, z1 = qq d Define the Lyapunov function: where c1∈R n×n is a positive definite symmetric constant matrix; definition: Taking the derivative of the Lyapunov function, we get: If z2=0, then is the quadratic function of the system's control output error z1, when The next step of design is required; Step 52: Define the sliding surface function: s = k1z1 + z2 Where: k1∈R n×n is a symmetric positive definite constant matrix; Define the Lyapunov function: Taking the derivative of the Lyapunov function, we get: Based on the previous backstepping sliding mode controller design, an adaptive algorithm is used to adjust the unobservable part of the external disturbance signal. Make an estimate; define For the unobservable part The estimated value of is defined as: Define the Lyapunov function: Among them: γ is a positive constant; after taking the derivative of the Lyapunov function, we can get: The control law designed by the above formula is as follows: Where: u eq is equivalent control; u vs is the switching control; h∈R n×nis a symmetric, positive constant matrix; β is a positive constant; sgn is a sign function; The adaptive law can be taken as:
[0036] In step 6, for the entire closed-loop system, the Lyapunov function can be written as follows: Taking the derivative of the Lyapunov function, we get: The above formula can be written as follows: in:
[0037] In order to The matrix P needs to be a positive definite matrix to ensure that the robot system satisfies the Lyapunov stability criterion, thereby ensuring that the robot system has exponential asymptotic stability in a global sense.
[0038] In order to better illustrate the present invention, MATLAB simulation is used to verify the proposed control strategy. Under the MATLAB2022b environment, the present invention uses simulink to record the designed adaptive inversion sliding mode control method for the robotic arm based on disturbance observer and friction compensation as AISMC-DE-FC as an experimental group and apply it to a two-degree-of-freedom robotic arm, and the traditional adaptive inversion sliding mode control method for the robotic arm system that only uses a disturbance observer is recorded as AISMC-DE as a control group, and the same sliding mode controller parameters and disturbance observer parameters are selected for simulation test verification.
[0039] The simulation parameters are as follows
[0040] The nonlinear function to be designed in the disturbance observer is defined as: in: Table 1
[0041] In the above parameters: L1, L2 are the lengths of the robot arm links; m1, m2 are the link masses of joints 1 and 2; J1, J2 are the inertial forces on joints 1 and 2. In order to obtain the ideal tracking characteristics, the parameters of the disturbance observer and the sliding mode controller are continuously adjusted to finally obtain a set of optimal parameters. The specific values of the obtained physical parameters of the robot arm and the parameters of the sliding mode controller and the disturbance observer are shown in Table 1.
[0042] The initial joint angle of the robot system is q = [1.2, 1.2] T , the initial joint angular velocity is taken as The expected trajectory of joint 1 is: 11d =0.1sin(0.5t)+cos(0.5t), the expected trajectory of joint 2 is: x 12d =0.1sin(t)+cos(t). The friction parameters of the LuGre friction model are shown in Table 2: Table 2
[0043] Result description:
[0044] Figure 3 , Figure 4 and Figure 6 , Figure 7 The following are simulation diagrams of the position tracking and speed tracking of the two joints of the robotic arm under the experimental group and control group methods, respectively. It can be seen from the figure that compared with the control group, the two robotic arms in the experimental group of the present invention can track the desired trajectory in a shorter time, which reflects the advantage of fast tracking of the present invention.
[0045] Figure 5 , Figure 8 Schematic diagrams of simulation of position tracking error and speed tracking error of two joints of the robot arm under the experimental group and control group methods respectively. It can be seen from the figure that compared with the control group, the steady-state errors of the two joints of the robot arm in the experimental group of the present invention are smaller, which reflects the advantage of high tracking accuracy of the present invention.
[0046] Fig. 9 Schematic diagram of the comparison of the joint torque of the mechanical arm between the experimental group and the control group. It can be seen from the figure that compared with the control group, the control input torque of the experimental group in the present invention remains continuous and the chattering phenomenon is significantly reduced.
[0047] In summary, the control method designed by the present invention can enable the manipulator system to quickly track the desired trajectory. Since the joint friction that the disturbance observer cannot accurately estimate is taken into account, it can show stronger robustness against disturbances. The inverse sliding mode controller combined with the adaptive control design significantly improves the chattering phenomenon of the sliding mode control, and also ensures that the manipulator system has global asymptotic stability.
[0048] It should be noted that the parts not described in detail in the above embodiments are all prior art.
[0049] The above description is only a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as a preferred embodiment as above, it is not intended to limit the present invention. Any technician familiar with the profession can make some changes or modifications to equivalent embodiments of equivalent changes using the technical contents disclosed above without departing from the scope of the technical solution of the present invention. However, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention should be covered within the protection scope of the present invention.
Claims
1. A method for adaptive backstepping sliding mode control of a robotic arm. It is characterized in that The following steps are involved: Step 1: Establish an n-dimensional robotic arm dynamics model based on Lagrangian dynamics mechanism; Step 2: Considering that the disturbance observer cannot accurately describe the joint friction torque, the dynamic model of the manipulator is modified and the mathematical model of the dynamics of the manipulator is converted into a state space equation form; Step 3, using the LuGre model to describe the friction torque that the disturbance observer cannot accurately observe, and performing offline identification of the friction parameters; Step 4: Design a disturbance observer to observe and compensate for uncertain parameters and unknown disturbances; Step 5: Design a sliding mode controller by combining the adaptive strategy and the backstepping sliding mode control; Step 6: Use the Lyapunov stability criterion to analyze the stability of the robotic arm system.
2. The method for adaptive backstepping sliding mode control of a manipulator according to claim 1, It is characterized in that The dynamic model of the robotic arm in step 1 is as follows: Where: M(q)∈R n×n is the positive definite inertia matrix of the robotic arm system; is the centripetal force and Coriolis force matrix of the robotic arm system; G(q)∈R n is the gravity vector acting on the robot joint; u∈R n Control input torque for the robot system; is the total unknown disturbance of the robotic system, where τ d is the unknown time-varying disturbance, ΔM, ΔC, ΔG are the modeling errors of the robot system; q, are the actual angle, angular velocity and angular acceleration of the robot.
3. The method for adaptive backstepping sliding mode control of a manipulator according to claim 2, It is characterized in that The specific steps of modifying the mechanical arm dynamics model and converting the mechanical arm dynamics mathematical model into the state space equation form in step 2 are as follows: For the joint friction torque F, F will switch between positive and negative when the robot joint turns. The first-order derivative of the joint friction torque at this time is Considering the joint friction torque as part of the total disturbance f, at this stage The joint friction torque cannot be simply regarded as part of the total disturbance f; in order to accurately feedforward compensate the robot system, the robot dynamics model can be written as follows: Where: f 1 is the total disturbance except the joint friction torque; F is the joint friction torque; make Then the state space equation of the robotic arm system can be expressed as:
4. The method for adaptive backstepping sliding mode control of a manipulator according to claim 3, It is characterized in that The LuGre friction torque model in step 3 is as follows: Where: v represents the relative speed of the two objects; v s is the Stribeck speed effect; T c , T s are Coulomb friction and static friction respectively; σ 0 is the stiffness coefficient of the bristle; σ 1 is the bristle micro damping coefficient; σ 2 is the viscous damping coefficient; σ 0 , σ 1 is the dynamic friction parameter; g(v) is the velocity-related function; z is the average deformation of the bristle.
5. The method for adaptive backstepping sliding mode control of a manipulator according to claim 4, It is characterized in that In step 3, a constant speed tracking experiment is performed to collect the original friction data, and the friction parameters are identified using identification technology. When the robot arm joint moves at a constant speed, the friction torque can be described as follows: f v =g(v)sgnv+σ 2 v.
6. The method for adaptive backstepping sliding mode control of a manipulator according to claim 5, It is characterized in that In step 4, it is assumed that the total unknown disturbance f satisfies continuous differentiability and is bounded and the bound is unknown, that is, Based on the dynamic model in step 1, a disturbance observer is designed to observe the system disturbance; the disturbance observer is as follows: in: is the nonlinear function to be designed; is the friction torque of the joint after identification; is the gain of the disturbance observer, is the estimated value of f; After adopting the disturbance observer, the robotic arm system in step 2 becomes in: is the observation error of the disturbance observer; u BS is the sum of equivalent control and switching control.
7. The method for adaptive backstepping sliding mode control of a manipulator according to claim 6, It is characterized in that After the disturbance observer is introduced in step 5, the state space equation in step 2 is written as: Combined with the relevant theories of sliding mode control, the adaptive backstepping sliding mode controller for the manipulator is designed in two steps.
8. The method for adaptive backstepping sliding mode control of a manipulator according to claim 7, It is characterized in that In step 5, the related theories of sliding mode control are combined to design the adaptive backstepping sliding mode controller of the manipulator in two steps. The specific steps are as follows: Step 51: Set q d Give the position command to the robot arm and define z 1 is the control output error of the system, that is, z 1 =qq d Define the Lyapunov function: where c 1 ∈R n×n is a positive definite symmetric constant matrix; definition: Taking the derivative of the Lyapunov function, we get: If z 2 =0, then is the control output error z of the system 1 A quadratic function, when Then the next step of design is required; Step 52, define the sliding surface function: s = k 1 z 1 +z 2 Where: k 1 ∈R n×n is a symmetric positive definite constant matrix; Define the Lyapunov function: Taking the derivative of the Lyapunov function, we get: Based on the previous backstepping sliding mode controller design, an adaptive algorithm is used to adjust the unobservable part of the external disturbance signal. Make an estimate; define For the unobservable part The estimated value of is defined as: Define the Lyapunov function: Among them: γ is a positive constant; after taking the derivative of the Lyapunov function, we can get: The control law designed by the above formula is as follows: Where: u eq is equivalent control; u vs is the switching control; h∈R n×n is a symmetric, positive constant matrix; β is a positive constant; sgn is a sign function; The adaptive law can be taken as:
9. The method for adaptive backstepping sliding mode control of a manipulator according to claim 8, Features: In step 6, for the entire closed-loop system, the Lyapunov function can be written as follows: Taking the derivative of the Lyapunov function, we get: The above formula can be written as follows: in:
10. The method for adaptive backstepping sliding mode control of a manipulator according to claim 9, It is characterized in that In order to The matrix P needs to be a positive definite matrix to ensure that the robotic arm system meets the Lyapunov stability criterion.
Citation Information
Cited By
Improved LuGre friction model-based self-adaptive robust control method for electric actuating mechanism
CN119781288A
Adaptive robust control method for electric actuator based on improved LuGre friction model
CN119781288B
Mechanical arm disturbance observation method and device and nonvolatile storage medium
CN119897863A
