Double-pneumatic artificial muscle control method and system with overshoot constraint

Abstract

The invention provides a double-pneumatic artificial muscle control method and system with overshoot constraint, and relates to the technical field of robot control. Comprising the following steps: defining error signals of a forearm muscle length, an upper arm muscle length, a shoulder joint angle and an elbow joint angle based on a dynamic model of a double-pneumatic artificial muscle system, and constructing an error signal convergence constraint term and an overshoot range constraint term; determining an energy function of the double-pneumatic artificial muscle system; based on a kinetic model and an energy function of the double-pneumatic artificial muscle system, considering the error signal convergence constraint term and the overshoot range constraint term, and designing a controller to control the forearm control input air pressure and the upper arm control input air pressure of the double-pneumatic artificial muscle system. According to the method, the error signal convergence constraint term and the overshoot range constraint term are considered, the overshoot of the system state variable is controlled within the safety constraint range, the overshoot phenomenon is avoided, and the safety performance of the system is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of robot control, and in particular relates to a dual-pneumatic artificial muscle control method and system with overshoot constraint. Background Art

[0002] In recent years, robots driven by PAMs (Pneumatic Artificial Muscles) have been widely studied in human-robot interaction scenarios due to their inherent safety and flexibility. The main structure of a PAM consists solely of a flexible, hollow rubber tube tightly wrapped with a braided sheath to achieve circumferential restraint. The tube's ends are sealed with metal joints, allowing compressed air to enter the muscle only through pre-defined vents at the joints. While conventional cylinder actuators offer a degree of flexibility due to their pneumatic drive characteristics, they are still inherently rigid due to the presence of sliding components such as pistons. PAMs, on the other hand, exhibit superior flexibility and safety, as they lack rigid components such as pistons. Their operating principle is as follows: During the intake phase, an air compressor (i.e., the air source) delivers high-pressure gas to the PAM through a valve via an air pipe. The pressure differential between the internal pressure of the rubber tube and the external atmospheric pressure causes the PAM to undergo radial expansion and axial contraction. During the exhaust phase, the PAM undergoes a reverse deformation process, with radial contraction and axial expansion, achieving its cyclical motion pattern.

[0003] PAMs are typically made of flexible materials and operated via gas actuation within a flexible tubular structure. Externally pressurized PAMs exhibit muscle contraction to reduce the actuation length, resulting in biomimetic muscle contraction motion, enhancing their applicability for assisted locomotion and assisted rehabilitation. In the early stages of research, researchers utilized adaptive and neural network approaches to control the position of a single PAM. Specifically, to achieve adequate tracking performance, an adaptive control method for a single PAM system was introduced. To improve the resilience of a single PAM system and mitigate the effects of sensor noise, a Kriging median compensator was proposed for modeling and correcting asymmetric hysteresis. By employing a backpropagation neural network-proportional-integral-derivative (BPNN-PID) algorithm in both the inner and outer control loops, the developed adaptive controller achieved significantly improved tracking accuracy compared to a traditional PID controller. This extensive research on single PAMs provides strong theoretical support for the joint design and development of PAMs.

[0004] However, individual PAMs can only stretch or contract in one direction, making them inflexible when performing complex, multi-dimensional motions. In real-world production and life, the use of single-dimensional PAMs is more limited. To accomplish more complex tasks and overcome the limitations of single PAM flexibility, various combinations of multiple PAMs are often used. Multi-PAMs offer high flexibility and relatively smooth motion, more accurately replicating the kinematic characteristics of human muscles. Consequently, they have found widespread application in biomimetic, medical, and rehabilitation fields. For example, by placing multiple PAMs at different joints in a robot, multi-directional motion control comparable in functionality to the human muscle system can be achieved. This enables robots to perform increasingly complex tasks such as grasping, manipulation, and posture adjustment, all of which significantly improve their performance. Furthermore, the widespread application of PAM-driven robots in fields such as chemical manufacturing, biomimetics, and off-road assistance has greatly expanded their scope and versatility. This widespread development in various fields has been accompanied by the flexible application of a variety of methods. For example, methods such as sliding mode control, adaptive control, and neural network control can be used to manage uncertainty in PAM systems, improving their robustness and control performance. In the area of ​​dual-PAM actuation control, Wang et al. equipped a dual-PAM-driven robot with an adaptive fuzzy backpropagation control system to assist patients in arm extension movements. Zhao et al. proposed an adaptive control law to address the input dead zone problem in dynamic nonlinear systems and further applied it to the angle tracking control of a PAM, effectively improving the trajectory tracking accuracy of the PAM. Wang et al. designed a mechanical model based on an artificial neural network and constructed a strain energy function for the PAM, verifying the accuracy of the proposed model.

[0005] As more and more research is conducted on PAM, the challenges of system performance optimization are becoming greater and greater, and there are still some practical problems that need to be solved urgently.

[0006] (1) A few control methods, such as adaptive control and neural network control, cannot guarantee that the controller can reliably maintain the specified range. However, in actual operation, the arm movement angle range and muscle contraction length are generally limited, which may cause related overshoot and cause unknown safety hazards to the system.

[0007] (2) When the denominator of the controller is equal to zero, a singularity problem will inevitably occur. The singularity problem may cause energy loss and unknown safety hazards, and it will also increase the calculation difficulty of the control input part.

[0008] (3) Linearization is usually only applicable to small changes in nonlinear systems near the operating point because it cannot accurately handle model uncertainties under reasonable assumptions. This means that the linearized model may be inaccurate in areas far from the operating point. Summary of the Invention

[0009] In order to overcome the shortcomings of the above-mentioned prior art, the present invention provides a dual-pneumatic artificial muscle control method and system with overshoot constraints, taking into account the error signal convergence constraint terms and overshoot range constraint terms, and controlling the overshoot of the system state variables within the safety constraint range to avoid possible dangers, ensure the effective operation of the controller, avoid overshoot phenomena, and improve the safety performance of the system.

[0010] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: A first aspect of the present invention provides a dual pneumatic artificial muscle control method with overshoot constraint.

[0011] A dual pneumatic artificial muscle control method with overshoot constraint comprises the following steps: Establish a dynamic model of a dual-pneumatic artificial muscle system; Based on the dynamic model of the dual-pneumatic artificial muscle system, error signals including forearm muscle length, upper arm muscle length, shoulder joint angle, and elbow joint angle are defined, and error signal convergence constraints and overshoot range constraints are constructed. Determine the energy function of the dual pneumatic artificial muscle system; Based on the dynamic model and energy function of the dual pneumatic artificial muscle system, the Lyapunov function is selected, and the control parameters are chosen according to the Lyapunov stability theory to ensure that the system meets the above-mentioned error signal convergence constraints and overshoot range constraints. A controller is designed to control the forearm control input pressure and the upper arm control input pressure of the dual pneumatic artificial muscle system.

[0012] A second aspect of the present invention provides a dual-pneumatic artificial muscle control system with overshoot constraint.

[0013] A dual-pneumatic artificial muscle control system with overshoot constraint includes: The dynamic model building module is configured to: build a dynamic model of the dual pneumatic artificial muscle system; The constraint item construction module is configured to: define error signals including forearm muscle length, upper arm muscle length, shoulder joint angle, and elbow joint angle based on a dynamic model of the dual-pneumatic artificial muscle system, and construct error signal convergence constraint items and overshoot range constraint items; The energy function determination module is configured to: determine the energy function of the dual pneumatic artificial muscle system; The controller design module is configured as follows: based on the dynamic model and energy function of the dual pneumatic artificial muscle system, a Lyapunov function is selected, and control parameters are selected according to the Lyapunov stability theory to ensure that the system meets the above-mentioned error signal convergence constraints and overshoot range constraints, and a controller is designed to control the forearm control input pressure and the upper arm control input pressure of the dual pneumatic artificial muscle system. A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of the dual-pneumatic artificial muscle control method with overshoot constraint as described in the first aspect of the present invention.

[0014] The fourth aspect of the present invention provides an electronic device, comprising a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, the steps of the dual pneumatic artificial muscle control method with overshoot constraint as described in the first aspect of the present invention are implemented.

[0015] One or more of the above technical solutions have the following beneficial effects: This paper discloses a dual-pneumatic artificial muscle control method and system with overshoot constraints. By appropriately varying the barrier function, two constraints are carefully constructed: an error signal convergence constraint and an overshoot range constraint. By controlling the overshoot of the system state variables within a safe range, potential hazards are avoided and the effective operation of the controller is ensured.

[0016] The method of the present invention ensures the denominator of the controller by reasonably designing the specific forms of items A and B, that is, , never reaches zero, preventing the controller from losing its control effectiveness and effectively resolving the singularity problem. The system was then subjected to in-depth theoretical analysis and verification using the Lyapunov method. This not only protected the system's normal operation but also prevented the negative impact of singularities.

[0017] Advantages of additional aspects of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0019] Figure 1 This is a schematic diagram of the dual PAM system model of Example 1.

[0020] Figure 2 For Example 1 Experiment 1 Positioning results.

[0021] Figure 3 For Example 1 Experiment 1 Positioning results.

[0022] Figure 4 For Example 1 Experiment 1 Positioning results.

[0023] Figure 5 For Example 1 Experiment 1 Positioning results.

[0024] Figure 6 Switching in Experiment 2 of Example 1 Positioning results.

[0025] Figure 7 Switching in Experiment 2 of Example 1 Positioning results.

[0026] Figure 8 For Example 1 Experiment 3 Robustness test results.

[0027] Figure 9 For Example 1 Experiment 3 Robustness test results.

[0028] Figure 10 This is a flow chart of the method of embodiment 1. DETAILED DESCRIPTION

[0029] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used in this embodiment have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0030] It should be noted that the terms used herein are for describing particular embodiments only and are not intended to limit the exemplary embodiments according to the present invention.

[0031] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.

[0032] Example 1 like Figure 10 As shown, this embodiment discloses a dual pneumatic artificial muscle control method with overshoot constraint.

[0033] A dual pneumatic artificial muscle control method with overshoot constraint comprises the following steps: Establish a dynamic model of a dual-pneumatic artificial muscle system; Based on the dynamic model of the dual-pneumatic artificial muscle system, error signals including forearm muscle length, upper arm muscle length, shoulder joint angle, and elbow joint angle are defined, and error signal convergence constraints and overshoot range constraints are constructed. Determine the energy function of the dual pneumatic artificial muscle system; Based on the dynamic model and energy function of the dual pneumatic artificial muscle system, the Lyapunov function is selected, and the control parameters are chosen according to the Lyapunov stability theory to ensure that the system meets the above-mentioned error signal convergence constraints and overshoot range constraints. A controller is designed to control the forearm control input pressure and the upper arm control input pressure of the dual pneumatic artificial muscle system.

[0034] This embodiment proposes a nonlinear robust controller for the dual pneumatic artificial muscle model, which takes into account overshoot constraints and singularity problems and has precise positioning performance and excellent robustness.

[0035] Furthermore, in order to enable the dual-PAM system to continue to respond accurately and stably in the face of uncertainty and interference, this embodiment carefully constructs a function to combine the system dynamics model with the carefully designed By properly selecting The control accuracy and response speed of the control system input are comprehensively demonstrated by Lyapunov, the dynamic model is coupled, and the , making the analysis of system stability issues easier.

[0036] 1 Dynamic model of dual pneumatic artificial muscles Figure 1 The dual pneumatic artificial muscle system model is shown in Table 1. The initial length of the dual pneumatic artificial muscle is The robot is initially suspended vertically on a horizontal plane. The lengths of the springs, muscles, and arm bones are the same, i.e., the initial lengths of the two springs are , the distance between the arm bone and the PAM (or spring) satisfy .

[0037] Define the auxiliary angle and auxiliary length as , where the auxiliary length .

[0038] The following formula can be used to express the upper arm / forearm muscle length and spring length :

[0039]

[0040] (1) joint angle It can be expressed as follows: (2) The derivative of (2) can be expressed as: (3) (4) Table 1. Parameters and variables

[0041] This example is based on the dual PAM kinetic model in the literature

[37] , namely ; ,in represents the inertia matrix; Represents the matrix element of the inertia matrix, and its specific expression is as follows:

[0042] ; ,in represents the Coriolis force matrix; , It represents the matrix element of the Coriolis force matrix, and its specific expression is as follows: ,

[0043] ,in represents the gravity vector, Represents the non-zero component of the gravity vector, and its specific expression is as follows , Represents the system state quantity, Represents the system control input.

[0044] ,

[0045]

[0046] The dynamic model is expressed as follows: (5) (6) (7) (8) in, 、 、 、 、 、 、 、 、 、 、 These are auxiliary variables defined to facilitate writing and simplify the dynamic model.

[0047] is the unit correlation coefficient of upper arm PAM contraction force, is the unit correlation coefficient of forearm PAM contraction force, is the upper arm PAM damping unit correlation coefficient, is the forearm PAM damping unit correlation coefficient, is the unit correlation coefficient of the upper arm PAM second-order spring, is the unit correlation coefficient of the second-order spring of the forearm PAM. To facilitate writing and simplify the dynamic model, the following auxiliary variables are defined:

[0048] In order to further analyze the singularity problem, we define : (9) in Respectively upper limit.

[0049] The error signal is defined as: (10) in, 、 、 、 are the length and angle error signals of the upper arm and forearm muscles, respectively.

[0050] Based on (10), the derivative of the error signal is: (11) 2.2 Control Objectives This embodiment proposes a nonlinear robust controller to achieve the following goals: the PAM state quantity is always kept within the allowable range and accurately reaches the desired position. The stability and robustness of the system are demonstrated through comprehensive theoretical analysis. Considering (2), the optimal angles of the shoulder joint and elbow joint can be expressed as follows: (12) The control objective of this embodiment is to establish a robust controller by combining the target position determined by the above calculations with the changed system dynamics to achieve the following points.

[0051] (1) The state variables converge to the expected values, so that the lengths and angles of the upper arm and forearm muscles satisfy:

[0052]

[0053]

[0054] (13) in, represent the expected values ​​for the forearm and upper arm muscles and the shoulder and elbow joint angles, respectively.

[0055] (2) Taking into account factors such as limited working area and safety considerations, the system state variables are always kept within the following range: (14) in, Represents state variables The maximum allowed overshoot.

[0056] III. Main achievements 3.1 Controller Design The energy of the two-link robot driven by PAM can be expressed as: ; in, ; ; Substitute into the previous dynamic model arrive (5.10) can be deduced: ;

[0057] The energy function of the dual PAM system is: (15) right Find the derivative with respect to time: (16) Based on the above study of the control objectives, the following controller is designed to avoid the singularity problem and suppress the system overshoot problem.

[0058] (17) in, is the positive gain that needs to be adjusted, For the auxiliary function that will be given later, 、 Respectively represent the control component composition in controller P1P2; 、 represents the stability term designed to ensure Lyapunov stability; and (18) (19) 3.2 Stability Analysis Theorem 1: For the PAM system, the proposed nonlinear robust controller (17) can not only guarantee the positioning performance of the system but also satisfy the overshoot constraints (13) and (14), that is, (20) Proof: Choosing a Lyapunov candidate function (twenty one) Can get Derivative with respect to time: (twenty two) Substitute (22) into (16) (twenty three) However, the singularity problem of controller design is challenging, so another candidate Lyapunov function is developed, namely (twenty four) Can get Derivative with respect to time: (25) in,

[0059] (26) and is a positive gain, and the sum of (23) and (25) gives the total Lyapunov function : (27) The derivative of (27) is: (28) Will Substituting (17) into (17) gives (29) Therefore, it can be concluded that the system is asymptotically stable, that is, (30) Substituting the designed controller (17) into (28) yields (31) The following four properties can be proved by (31): (1) Assumption ,but , which contradicts (30), so, The overshoot value will not exceed the maximum overshoot value setting value : (32) (2) Assumptions ,but , which contradicts (30), so, The overshoot value will not exceed the maximum overshoot value setting value : (33) (3) According to (30), the following conditions can be proved: (34) (4) In order to verify whether the constraint (9) always holds, assume that exist The limit is about to be exceeded due to is continuous, if There is a tendency to go beyond the given range, before which it must moment, approaching and reaching the boundary value, they must first approach the boundary value 0 or , it can be deduced that exist When (35) The conclusion drawn from (35) contradicts (31). Always within the given range.

[0060] The above research shows that the proposed control method can ensure the constraint conditions (9) in the entire control process and Further analysis of the balance point of the PAM-driven two-link robot shows that Only when the positioning error converges to the equilibrium point, that is, .

[0061] According to (30) and (17), we can conclude that (36) Combining (5)-(8) with (31), we can obtain (37) The time integral is: (38) Combining (36), (37) and Barbalat's lemma, we can derive

[40] : (39) Noting the relationship between (39) and (3), we can obtain: (40) therefore (41) in, where is a constant to be determined.

[0062] By rewriting the kinetic models (5) and (6), we can obtain the variables Approaches zero. Through model transformation, (5) can be rewritten as follows (42) in

[0063]

[0064] Due to (39), considering (31) and (39), we can get , Using (39) and the extended Barbalat lemma, we can infer that: (43) Similarly, combining (6), (31), (39), and (36) we can obtain: (44) (7) can be written as follows: (45) Furthermore, combining (39) and (43) we can obtain: (46) Applying the same method, based on (8), (39), and (44), we can derive (47) Substitute (39), (43), and (44) into the kinetic models (5) and (6) (48) (49) Combining the specific form of the controller (17) and the conclusions drawn from (39), (43), and (44), we can simplify (48) and (49) to obtain: (50) (51) Substitute (22) into the controller , and combined with (50) and (51) we can get (52) According to (48) and (49), we can draw the final conclusion: (53) Applying the same method, we can get , we can deduce from (10) (54) Substituting (41) into (12), the expected joint angle between the elbow and shoulder can be described as (55) Obviously, will also converge to zero. Therefore, we can draw the following conclusions: (56) According to conclusion (56), Theorem 1 is proved.

[0065] like Figure 10As shown, the energy function in this embodiment is designed based on the dynamic model, including parameters and variables in the dynamic model; Lyapunov stability analysis is performed, that is, the convergence of each state quantity of the system is analyzed; P1 and P2 in the dynamic model, namely the energy functions P1 and P2, namely the controller input P1 and P2, and the controller design is a bridge connecting the controller with the Lyapunov candidate function, the energy dynamics model, and the state quantity, and the Lyapunov function is used to further verify whether the designed controller is stable and whether the state quantity converges.

[0066] 4. Experimental Results 4.1 Software and Hardware Design of Serial Dual Pneumatic Artificial Muscles This example uses a PAM robotic arm experimental platform to conduct multiple experiments. Extensive testing was performed to confirm the positioning performance, robustness, and safety of the proposed control method, comparing it to a dynamic neural network-PID (DNN-PID) control method and current traditional PID control methods. The PAM robotic arm experimental platform configuration consists of two main components: (1) Hardware configuration: An air compressor (TYW-700) generates compressed air at 0-6 bar. To ensure the pneumatic channel, the compressed air enters a proportional pressure regulator (Festo VPPM-6L-L-1-G18-0 L 6 HV 1 P-C1) and then enters the upper arm muscle (Festo DMSP-20-200 N-RM-CM-DN) and forearm muscle (Festo DMSP-20-160 NRM-CM-DNPAM). When the internal air pressure increases, the PAM contracts, pushing the robot to rotate and generate joint angles.

[0067] (2) Software Configuration: The motion control panel uses a Googletech GTS-800-PV, the incremental encoder uses a Tamagawa OIH-48, and the host computer measures the signal via MATLAB / Simulink. During the transmission process, the signal is sent from the host computer to the air valve via the motion control panel. To regulate and control the gas flow, the motion control panel opens the Festo pressure regulator based on the actual control signal calculated by the host computer every two milliseconds. The incremental encoder measures the angle data synchronously with the host computer by measuring the joint angle.

[0068] The sampling time for data processing is set to 2 ms, and the sampling period is less than 10% of the step response settling time, which is sufficient to fully identify the dynamic characteristics of complex nonlinear systems.

[0069] 4.2 Experimental Results and Analysis To ensure the accuracy and reliability of the experiment, the controller gain and system parameters are determined. The system parameters are: (57) The parameters of the pneumatic artificial muscle are as follows: (58) The controller gains are set to: (59) The PID control method and DNN-PID controller are selected as comparison methods to evaluate the control performance.

[0070] (60) Autotuning gain matrix and update rate The detailed expression of can be found in Ho PH A. Online tuning gain scheduling MIMO neural PID control of the 2-axes pneumatic artificial muscle (PAM) robot arm [J]. Expert Systemswith Applications, 2010, 37(9): 6547-6560.

[0071] Three sets of experimental verifications were carried out on a tandem dual-pneumatic experimental platform to evaluate the control performance.

[0072] Experiment 1: Positioning Test In order to fairly verify the localization performance of the proposed method, we set the dual PAM length and the joint angle between muscles as two sets of localization targets. The PID control gain is selected as: (61) The DNN-PID control gains are selected as: (62) Case 1: The joint angles between muscles and the optimal muscle lengths are determined as: .

[0073] Case 2: The joint angles between muscles and the optimal muscle lengths are determined as: .

[0074] The results of Experiment 1 are as follows Figure 2 、 Figure 3 、 Figure 4 、 Figure 5 、 Figure 6 As shown. Figure 3-4 As shown, using the suggested method, muscle length and joint angles It quickly converges to the steady-state value within 3.2 seconds with the minimum steady-state error. In contrast, DNN-PID control and traditional PID control require nearly 4 seconds and 7 seconds to converge to the equilibrium value, respectively. Figure 5-6 As shown in Figure 3, when the positioning position is changed and the three methods are compared again, the proposed method can still reach the desired position within 3.4 seconds with good positioning performance.

[0075] Experiment 2: Switching position positioning test In real life, PAM is often used in applications such as medical robots, which usually require frequent position changes. Therefore, the controller based on dual pneumatic muscle design needs to be able to quickly adapt to various positioning tasks while maintaining accuracy and stability.

[0076] In this experiment, the length and joint angle of the dual pneumatic muscle system were initially set to At the 20th second, the system's positioning information is switched to .

[0077] The experimental results are as follows Figure 7-Figure 8 As shown in the figure, compared to the traditional PID control method, which takes 7 seconds to reach the ideal position, the method proposed in this example reaches the ideal position faster at the beginning of the experiment. At the same time, there is a large oscillation amplitude when switching the target position. Although the convergence speed of DNN-PID control is similar to that of the proposed method, there is still large oscillation when switching the target position. Therefore, the proposed robust control can effectively suppress overshoot during the switching process.

[0078] Experiment 3: Robustness Test To further verify whether the proposed technique is more robust, we applied external interference to the dual PAM. The same amplitude was applied at 10 seconds, 20 seconds, and 30 seconds. The experimental results are shown in Figure 2. Figure 9-10 shown.

[0079] Experimental results of the proposed method are compared with those of other methods. The proposed method can quickly and smoothly eliminate residual oscillations caused by disturbances and effectively minimize their negative effects. Furthermore, the overshoot of the proposed method under disturbances consistently satisfies the given constraints.

[0080] However, the traditional PID comparison method exhibits relatively large oscillation amplitudes and fails to reach the desired position after a given disturbance ends, requiring at least three seconds to return to a stable state. Furthermore, under external disturbances, overshoot significantly exceeds the set range. The DNN-PID control method exhibits some overshoot, and the residual oscillation amplitude after the disturbance is significantly greater than that of the proposed method. Therefore, in addition to reducing overshoot and eliminating residual oscillations, the robust controller described above also provides sufficient suppression of external disturbances.

[0081] Specifically, this embodiment minimizes the need for training large datasets and achieves precise positioning of artificial muscle-related state variables. A comprehensive Lyapunov stability study is also provided without the use of linearization. Finally, hardware testing is performed on a dual PAMs platform. A robust controller created based on system overshoot limits ensures rapid convergence of the state signal. The signal has strong anti-interference capabilities against interfering signals and can successfully reach the desired position even when encountering positioning switching issues. The practicality and safety of the proposed method are further confirmed.

[0082] Example 2 This embodiment discloses a dual-pneumatic artificial muscle control system with overshoot constraint.

[0083] A dual-pneumatic artificial muscle control system with overshoot constraint includes: The dynamic model building module is configured to: build a dynamic model of the dual pneumatic artificial muscle system; The constraint item construction module is configured to: define error signals including forearm muscle length, upper arm muscle length, shoulder joint angle, and elbow joint angle based on a dynamic model of the dual-pneumatic artificial muscle system, and construct error signal convergence constraint items and overshoot range constraint items; The energy function determination module is configured to: determine the energy function of the dual pneumatic artificial muscle system; The controller design module is configured as follows: based on the dynamic model and energy function of the dual pneumatic artificial muscle system, a Lyapunov function is selected, and control parameters are selected according to the Lyapunov stability theory to ensure that the system meets the above-mentioned error signal convergence constraints and overshoot range constraints, and a controller is designed to control the forearm control input pressure and the upper arm control input pressure of the dual pneumatic artificial muscle system.

[0084] Example 3 The purpose of this embodiment is to provide a computer-readable storage medium.

[0085] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the dual-pneumatic artificial muscle control method with overshoot constraint as described in Example 1 of the present disclosure.

[0086] Example 4 The purpose of this embodiment is to provide an electronic device.

[0087] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, the steps of the dual pneumatic artificial muscle control method with overshoot constraint as described in Example 1 of the present disclosure are implemented.

[0088] The steps involved in the apparatuses of Examples 2, 3, and 4 above correspond to those of Method Example 1. For detailed implementations, please refer to the relevant description of Example 1. The term "computer-readable storage medium" should be understood to mean a single medium or multiple media containing one or more instruction sets; it should also be understood to include any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and causing the processor to perform any method of the present invention.

[0089] Those skilled in the art will appreciate that the modules or steps of the present invention described above can be implemented using a general-purpose computer device. Alternatively, they can be implemented using program code executable by a computing device, which can then be stored in a storage device and executed by the computing device. Alternatively, they can be fabricated into separate integrated circuit modules, or multiple modules or steps can be fabricated into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0090] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. A dual pneumatic artificial muscle control method with overshoot constraint, characterized in that: The following steps are involved: Establish a dynamic model of a dual-pneumatic artificial muscle system; Based on the dynamic model of the dual-pneumatic artificial muscle system, error signals including forearm muscle length, upper arm muscle length, shoulder joint angle, and elbow joint angle are defined, and error signal convergence constraints and overshoot range constraints are constructed. Determine the energy function of the dual pneumatic artificial muscle system; Based on the dynamic model and energy function of the dual pneumatic artificial muscle system, the Lyapunov function is selected, and the control parameters are chosen according to the Lyapunov stability theory to ensure that the system meets the above-mentioned error signal convergence constraints and overshoot range constraints. A controller is designed to control the forearm control input pressure and the upper arm control input pressure of the dual pneumatic artificial muscle system.

2. The dual pneumatic artificial muscle control method with overshoot constraint according to claim 1, characterized in that: The dynamic model of the dual pneumatic artificial muscle system is specifically as follows: ； ； ； ； in, and Indicates the length of the upper arm and forearm muscles; represents the system control input, and Respectively represent the control components; Inertia matrix ; Coriolis force matrix ; Gravity vector ; 、 、 、 、 、 Represents the matrix elements of the inertia matrix; 、 、 、 、 、 represents the matrix elements of the Coriolis force matrix; 、 The elements representing the gravity vector; 、 、 、 、 、 、 、 、 、 These are auxiliary variables defined to facilitate writing and simplify the dynamic model; Represents the non-zero components of the gravity vector.

3. The dual pneumatic artificial muscle control method with overshoot constraint according to claim 2, characterized in that: The error signal is defined as: ； in, represent the expected values ​​of upper arm muscles and forearm muscles, and shoulder and elbow joint angles, respectively; and Indicates the length of upper arm muscles and forearm muscles; 、 Indicates the shoulder and elbow joint angles; 、 、 、 They are the error signals of upper arm muscle length, forearm muscle length, shoulder joint and elbow joint angle respectively.

4. The dual pneumatic artificial muscle control method with overshoot constraint according to claim 3, characterized in that: The error signal convergence constraint term is specifically: ； The overshoot range constraint item is specifically: ； in, Represents state variables The maximum allowed overshoot.

5. The dual pneumatic artificial muscle control method with overshoot constraint according to claim 3, characterized in that: The controller is specifically: ； ； in, is a positive gain, is an auxiliary function, and ； ； in, 、 Respectively represent the control component composition in controller P1P2; 、 represents the stability term designed to ensure Lyapunov stability; Represents the system state quantity, Represents the system control input.

6. The dual pneumatic artificial muscle control method with overshoot constraint according to claim 2, characterized in that: Also included is the definition of the coefficients of the forearm control input air pressure and the upper arm control input air pressure of the dual pneumatic artificial muscle system in the dynamic model of the dual pneumatic artificial muscle system. : ； in Respectively upper limit.

7. The dual pneumatic artificial muscle control method with overshoot constraint according to claim 1, characterized in that: It also includes the application of the Lyapunov method to verify the control performance of the designed dual-pneumatic artificial muscle system controller.

8. A dual-pneumatic artificial muscle control system with overshoot constraint, characterized in that: include: The dynamic model building module is configured to: build a dynamic model of the dual pneumatic artificial muscle system; The constraint item construction module is configured to: define error signals including forearm muscle length, upper arm muscle length, shoulder joint angle, and elbow joint angle based on a dynamic model of the dual-pneumatic artificial muscle system, and construct error signal convergence constraint items and overshoot range constraint items; The energy function determination module is configured to: determine the energy function of the dual pneumatic artificial muscle system; The controller design module is configured as follows: based on the dynamic model and energy function of the dual pneumatic artificial muscle system, a Lyapunov function is selected, and control parameters are selected according to the Lyapunov stability theory to ensure that the system meets the above-mentioned error signal convergence constraints and overshoot range constraints, and a controller is designed to control the forearm control input pressure and the upper arm control input pressure of the dual pneumatic artificial muscle system.

9. A computer-readable storage medium having a program stored thereon, characterized in that: When the program is executed by a processor, the steps of the dual pneumatic artificial muscle control method with overshoot constraint as described in any one of claims 1 to 7 are implemented.

10. An electronic device comprising a memory, a processor, and a program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the dual pneumatic artificial muscle control method with overshoot constraint as described in any one of claims 1 to 7 are implemented.

Images