A force-stiffness closed-loop control method and system for a variable-stiffness actuator based on force feedback

Abstract

The present invention belongs to the field of robot control, and discloses a force-stiffness closed-loop control method and system for a variable stiffness actuator based on force feedback, which includes a nonlinear PID control method, a stiffness closed-loop control method, and a parameter tuning method and process. The present invention can effectively improve the force control accuracy in highly nonlinear systems and complex disturbance environments through a nonlinear PID control method based on hyperbolic secant construction; the stiffness closed-loop control method based on force feedback can effectively adjust the system stiffness characteristics under different stiffness environments, thereby improving system stability and force control accuracy; the parameter tuning method of the control system based on heuristic algorithms and comprehensive objectives can effectively tune various parameters of the control system to ensure the stability of the system. This comprehensive technical feature makes the control method flexible, efficient, and adaptable in the field of robotics.

Description

Technical Field

[0001] The present invention belongs to but is not limited to the field of robot control, and in particular relates to a force-stiffness closed-loop control method and system for a variable stiffness actuator based on force feedback. Background Art

[0002] With the increasing demand for surface processing accuracy of complex curved parts, such as aircraft engine blades, precision molds, and complex medical devices, grinding and polishing technology is facing increasingly higher requirements. Traditional manual polishing methods, due to their reliance on manual operation, have problems such as high workload, poor repeatability, low processing efficiency, and difficulty in controlling precision. They can no longer meet the needs of modern high-precision processing. Therefore, automated robotic grinding and polishing systems have gradually become a key research and application direction. As the complexity of the processed shapes increases, research on variable stiffness actuators at the end of grinding and polishing robots has gradually increased. Among them, force-stiffness feedback control technology is considered to be the key to improving the processing accuracy of robotic grinding and polishing systems. However, existing robotic grinding and polishing systems still have many technical difficulties in force control and stiffness control that need to be solved.

[0003] Dynamic force-stiffness matching: During complex surface machining, the robot end-effector must maintain a stable contact force with the workpiece surface. However, due to the complex surface geometry and large variations in ambient stiffness, traditional stiffness control methods are unable to respond to these changes in real time, resulting in unstable contact force and, in turn, impacting polishing quality. Existing force control systems often rely on static parameter adjustments and lack adaptive mechanisms for dynamic ambient stiffness adjustments.

[0004] Large errors in environmental stiffness estimation: Current closed-loop stiffness control methods rely on the robotic grinding and polishing system to accurately estimate the environmental stiffness. However, due to the presence of numerous noise interferences and nonlinear factors in the machining environment, these errors in environmental stiffness estimation can significantly affect the accuracy of force feedback control. Furthermore, while variable-stiffness actuators can adjust their own stiffness within a certain range, they are unable to perceive and estimate changes in external stiffness in real time, limiting the performance of force-stiffness matching control.

[0005] Nonlinear response issues in control strategies: Existing linear and nonlinear PID control methods exhibit significant limitations in their control performance when processing complex curved surfaces, facing multiple sources of disturbance and nonlinear factors. To improve machining accuracy, robotic grinding and polishing systems require a control strategy that can adapt to complex environments and possess stronger nonlinear response capabilities to achieve precise adjustment of force feedback and stiffness control. Summary of the Invention

[0006] In view of the problems existing in the prior art, the present invention provides a force-stiffness closed-loop control method and system for a variable stiffness actuator based on force feedback.

[0007] The present invention is achieved by providing a force-stiffness closed-loop control method for a variable stiffness actuator based on force feedback, which includes a nonlinear PID control method, a stiffness closed-loop control method, and a parameter setting method and process.

[0008] Furthermore, the nonlinear PID control method is obtained by constructing a nonlinear gain function and multiplying the nonlinear gain function with a linear PID gain coefficient, which is expressed as:

[0009]

[0010] Where e is the error, which is the difference between the sensor measurement value and the expected value, K P Indicates the linear PID proportional coefficient, K I Indicates the linear PID integral coefficient, K D Indicates the linear PID differential coefficient, H P (e) represents the nonlinear gain function of the proportional coefficient and is related to the error, H I (e) represents the nonlinear gain function of the integral coefficient and is related to the error, H D (e) represents the nonlinear gain function of the differential coefficient and is related to the error, uc represents the output expectation of the controller, and t is time.

[0011] Furthermore, the nonlinear gain function includes two forms, A and B, both of which are constructed by hyperbolic tangent functions;

[0012] The expression of the A-type nonlinear gain function is H A (e) = 1 + h 11 (1-sech(h 12 e)), H A (e) is the value of the nonlinear gain function; h 11 and h 12 is the nonlinear parameter of the type I nonlinear gain function; e is the error, which is expressed as the difference between the sensor measurement value and the expected value;

[0013] Type A nonlinear gain function, when the feedback error e tends to 0, the nonlinear gain function tends to 1; when the feedback error e tends to infinity, the nonlinear gain function tends to 1+h 11 ;

[0014] The B-type nonlinear gain function is H B (e) is the value of the nonlinear gain function; h 21 and h 22 is the nonlinear parameter of the B-type nonlinear gain function; e is the error, which is expressed as the difference between the sensor measurement value and the expected value;

[0015] Type B nonlinear gain function, when the feedback error e tends to 0, the nonlinear gain function tends to 1; when the feedback error e tends to infinity, the nonlinear gain function tends to

[0016] By combining type A and type B nonlinear gain functions, a variety of nonlinear gain functions can be obtained; the nonlinear gain functions in the order of proportion, integration, and differentiation can be expressed as nine combinations: AAA, AAB, ABA, ABB, BAA, BAB, ABB, BBA, and BBB.

[0017] Furthermore, the stiffness closed-loop control method is based on a closed-loop stiffness feedback rate constructed by force feedback, tangent function, and exponential function, and changes the system stiffness by controlling the variable stiffness module through force feedback;

[0018] The closed-loop stiffness feedback rate discrete time domain expression:

[0019]

[0020] Among them, k vs (t) represents the expected stiffness at time t, k vs (t-1) represents the desired stiffness at time t, T is the discrete period of the control system, Δe represents the forward difference of the error e at time t, α1, α2, and α3 are the adjustment parameters of the stiffness control rate, and e is the error, which is expressed as the difference between the sensor measurement value and the expected value;

[0021] The closed-loop stiffness feedback rate is based on Evaluate the current rate of change of the force error; use a hyperbolic tangent function instead of a sign function to reduce chatter in stiffness control caused by force error noise interference; if the force error rate of change is positive, the stiffness is increased to reduce the damping ratio and response time, thereby enhancing the system's ability to quickly adjust to the force error. If the force error rate of change is negative, the stiffness is reduced to increase the damping ratio and improve system stability, thereby enhancing the system's anti-interference performance;

[0022] The magnitude of the stiffness change depends on the exponential decay function (1-exp(α3e f (t))); As the absolute error changes, larger force errors lead to larger adjustments, while smaller errors produce smoother changes.

[0023] Furthermore, the parameter setting method and process are as follows:

[0024] S1: Identified by the control system

[0025] S2: Linear parameter identification of nonlinear PID: K P , K I , K D

[0026] S3: Nonlinear gain function parameter identification of nonlinear PID: H P (e), H I (e), H D (e)

[0027] S4: Identification of stiffness control law parameters: α1, α2, and α3

[0028] The parameters of the controlled system in S1 can be identified by using the sweep frequency method or the pseudo-random sequence method to obtain the transfer function of the system;

[0029] The linear parameter identification in S2 can be carried out by the empirical accumulation method, the Ziegler-Nichols method, and the Cohen-Coon method;

[0030] The nonlinear gain function parameter identification in S3 is implemented by a heuristic optimization algorithm, such as a particle swarm algorithm or a genetic algorithm, but the algorithm optimization target is limited to the weighted sum of the output of the controlled system and the variance of the nonlinear PID control output as the objective function, which is expressed as follows:

[0031] Where α and β are weights, and J is the optimization target; The variance of the output of the controlled system is used to judge the stability of the system output; The variance of the nonlinear PID control output is used to evaluate the anti-disturbance capability of the nonlinear PID control system. When the controlled system is relatively stable and the external disturbance is relatively small, a larger α value can be set. When the external disturbance is relatively large, a larger β value can be set.

[0032] Furthermore, during the parameter tuning process of the heuristic optimization algorithm in S3, the output expected value of the controlled system is set to 0, a random disturbance is set in the output of the controlled system, and the stiffness control law is not applied;

[0033] The stiffness control law parameter identification in S4 is implemented by a heuristic optimization algorithm, such as a particle swarm algorithm or a genetic algorithm. However, the optimization goal of the algorithm is limited to the weighted sum of the output of the controlled system and the variance of the output of the control system as the objective function, which is expressed as follows:

[0034] Where α and β are weights, and J is the optimization target; The variance of the output of the controlled system is used to judge the stability of the system output; is the variance of the nonlinear PID control output, σ s 2 The stiffness control rate controls the output variance and is used to evaluate the anti-disturbance capability of the entire control system;

[0035] During the parameter tuning process of the heuristic optimization algorithm in S4, the output expected value of the controlled system is set to 0, and a random disturbance is set in the output of the controlled system.

[0036] Another object of the present invention is to provide a force-stiffness closed-loop control system for a variable stiffness actuator based on force feedback based on the force-stiffness closed-loop control method for a variable stiffness actuator based on force feedback, the system comprising:

[0037] The nonlinear PID module is obtained by constructing a nonlinear gain function and multiplying the nonlinear gain function with the linear PID gain coefficient;

[0038] The stiffness control rate module is based on the closed-loop stiffness feedback rate constructed by force feedback, tangent function, and exponential function. It controls the variable stiffness module through force feedback to change the system stiffness.

[0039] Parameter tuning module, identified by the control system; linear parameter identification of nonlinear PID: K P , K I , K D ; Nonlinear gain function parameter identification of nonlinear PID: H P (e), H I (e), H D (e); Identification of stiffness control law parameters: α1, α2 and α3.

[0040] Another object of the present invention is to provide a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the force-stiffness closed-loop control method of the variable stiffness actuator based on force feedback.

[0041] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to perform the steps of the force-stiffness closed-loop control method of a variable stiffness actuator based on force feedback.

[0042] Another object of the present invention is to provide an information data processing terminal, which is used to implement the variable stiffness actuator force-stiffness closed-loop control system based on force feedback.

[0043] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0044] 1. Adaptive force-stiffness adjustment: By combining nonlinear PID control based on a hyperbolic secant function with stiffness feedback, the system can adaptively respond to dynamic stiffness changes during complex surface machining, achieving highly precise force control. This approach effectively addresses the lag in force-stiffness matching response found in existing technologies, improving machining accuracy and efficiency.

[0045] 2. Enhanced Disturbance Immunity: This method uses heuristic optimization algorithms (such as particle swarm optimization or genetic algorithms) to tune control system parameters, ensuring optimal weighted variance between the system output and the nonlinear PID output, enabling precise response to complex disturbances and nonlinear factors. Furthermore, by combining multiple nonlinear gain functions, this method possesses strong adaptability and disturbance immunity.

[0046] 3. Smooth Stiffness Control: By designing a closed-loop stiffness feedback rate, this system can smoothly adjust stiffness in the presence of ambient noise, avoiding vibration caused by noise. Specifically, when the force error rate is positive, stiffness is rapidly increased to improve system response speed, while when the error rate is negative, the system reduces stiffness to enhance stability. This significantly improves the system's stability and compliance deficiencies found in existing technologies.

[0047] 4. Flexibility of Dual-Mode Nonlinear PID: The combination of Type A and Type B nonlinear gain functions provides a rich set of nonlinear adjustment methods (e.g., nine combinations, including AAA, AAB, and ABA). This allows for flexible system adjustments tailored to specific machining tasks, enhancing adaptability under diverse and complex working conditions. This is particularly true in fields such as aerospace, precision mold manufacturing, and medical devices. Its flexible parameter tuning and efficient closed-loop control approach not only improve the machining accuracy of single workpieces, but also significantly enhance the efficiency of batch machining of complex curved parts.

[0048] 5. Scalable Optimization Control: Using particle swarm optimization and genetic algorithms to tune PID parameters and stiffness control laws, the system can quickly find the optimal control parameter combination for different machining tasks and operating conditions, improving system response speed and control accuracy. Furthermore, in noisy environments, the system can appropriately enhance interference rejection by adjusting the variance weight, demonstrating significant adaptive adjustment advantages.

[0049] Second, the expected benefits and commercial value of the technical solution of the present invention after transformation are:

[0050] By introducing nonlinear PID control and stiffness closed-loop control methods, the accuracy of force-stiffness matching can be significantly improved, achieving stable contact force at the robot end during complex surface machining, reducing surface defects and machining errors. This is particularly suitable for surface machining of high-precision, complex curved parts such as aircraft engine blades and complex medical devices, helping to improve product performance and reliability.

[0051] The proposed control strategy can adapt to changes in rigidity in complex environments, effectively addressing multi-source disturbances and nonlinear factors, and reducing the impact of external environmental changes on machining quality. It demonstrates strong adaptability in machining workpieces of varying materials, shapes, and hardness, expanding the system's applicability. The system's high robustness and stability further reduce the incidence of equipment failures and improve overall economic benefits. This strategy addresses the increasingly stringent demands for machining quality in sectors such as aerospace, medical devices, and high-end molds, and promotes the industrialization and upgrading of robotic grinding and polishing system technology.

[0052] The technical solution of this invention fills a technological gap in the industry, both domestically and internationally. Compared with traditional static force control systems, this invention improves the system's ability to respond to external disturbances through real-time closed-loop force-stiffness control. Existing technologies often rely on linear PID control, which is unable to quickly respond to nonlinear disturbances and changes in environmental stiffness during complex surface machining. However, this invention, by combining nonlinear PID with a heuristic algorithm, enables the system to maintain high precision and efficiency even under complex machining conditions, thus filling a technological gap both domestically and internationally.

[0053] The technical solution of the present invention solves the technical problems that people have been eager to solve but have never been able to solve successfully:

[0054] 1. Real-time dynamic force-stiffness matching: In the grinding and polishing of complex curved parts, traditional systems are unable to adjust the end effector's stiffness and contact force in real time and stably, resulting in limited machining accuracy. This invention combines force feedback closed-loop control with a hyperbolic secant nonlinear PID algorithm to achieve dynamic adaptive adjustment of external stiffness, ensuring contact force stability during machining.

[0055] 2. Challenges in Estimating Environmental Stiffness: Due to noise interference and nonlinear factors, existing methods suffer from large errors in estimating environmental stiffness, which impacts control accuracy. The stiffness closed-loop feedback system proposed in this paper is based on the dynamic variation of force error and employs a hyperbolic secant function to reduce noise-induced control oscillation. This directly avoids estimating environmental stiffness and resolves this long-standing bottleneck.

[0056] 3. Balancing disturbance immunity and response speed: Traditional PID control methods struggle to balance rapid system response and stability in the face of multi-source disturbances. This paper, by constructing Type A and Type B hyperbolic secant nonlinear gain functions, implements a force error rate-driven stiffness regulation strategy, achieving a good balance between disturbance immunity and rapid response.

[0057] 4. Efficient Tuning of Multivariable Control Parameters: This invention uses heuristic optimization algorithms (such as particle swarm optimization or genetic algorithms) to tune multiple control system parameters, using weighted variance as the optimization target. This ensures high consistency between the system's control output and actual output. This tuning method overcomes the existing difficulty in simultaneously optimizing multi-parameter systems, resulting in a more adaptable and robust control system.

[0058] By introducing a closed-loop force-stiffness control method, this invention overcomes multiple technical biases of traditional technologies in robotic grinding and polishing systems, including insufficient stiffness adaptability in dynamic environments, low tolerance for noise interference, lack of nonlinear response capability, limited actuator stiffness adjustment range, and insufficient optimization capabilities in complex environments. By introducing a hyperbolic tangent nonlinear gain function and combining it with a heuristic algorithm for parameter tuning, the stiffness feedback rate design is optimized, achieving rapid adaptation to changes in external stiffness, robust suppression of noise interference, and coordinated optimization of system performance under multi-source disturbances. This significantly improves the force-stiffness matching performance and anti-interference capability in complex surface machining, providing innovative technical support for high-precision, high-efficiency robotic grinding and polishing.

[0059] Third, the technical solutions of the present invention solve the following problems:

[0060] In existing variable-stiffness actuator systems, the control system often faces problems such as inaccurate stiffness adjustment, slow response speed, system instability, and the inability to adaptively adjust under dynamic load conditions. These issues affect the performance of the actuator, especially in complex environments where high force and stiffness adjustment accuracy is required. Traditional PID control methods, which fail to account for nonlinear factors, are prone to error accumulation, delayed system response, and instability under load changes. In addition, traditional stiffness control methods do not fully utilize force feedback for real-time stiffness adjustment. As a result, in practical applications, the coordination between system stiffness and force output is not precise enough, making it impossible to achieve accurate force-stiffness closed-loop control.

[0061] Significant technical advancements of the present invention:

[0062] The force-stiffness closed-loop control method for a variable-stiffness actuator based on force feedback proposed in this paper significantly improves the performance of the variable-stiffness actuator by introducing a nonlinear PID control method, a stiffness closed-loop control method, and a parameter tuning method. Specific improvements are reflected in the following aspects:

[0063] 1. Introduction of nonlinear gain function: The present invention introduces a nonlinear gain function in PID control. Through the A-type and B-type gain functions of the hyperbolic tangent function, the gain coefficients of the proportional, integral, and differential can be dynamically adjusted, thereby optimizing the control accuracy. Especially in the case of large or small errors, the system response is smoother, avoiding the problems of overshoot and oscillation caused by traditional PID control when the error is large.

[0064] 2. Force Feedback and Stiffness Adjustment Closed-Loop Control: This invention's closed-loop stiffness control method utilizes force feedback and a closed-loop stiffness feedback rate constructed using a nonlinear function. This allows for real-time adjustment of actuator stiffness, closely aligned with force output, ensuring the stability and responsiveness of variable-stiffness actuators under varying load conditions. Through discrete-time stiffness feedback control, the system can more precisely adjust the desired stiffness and optimize the actuator's force-stiffness control characteristics.

[0065] 3. Multi-parameter adjustment control strategy: The present invention can flexibly adjust the response speed and stability of the system in practical applications by adjusting multiple parameters of the stiffness control rate, including proportional, integral and differential control coefficients, so that the system can adapt to changing load environments, thereby providing efficient and stable force and stiffness control in a wide range of application scenarios.

[0066] Through these technological advances, the present invention can provide a more accurate, rapid and stable force-stiffness closed-loop control solution, significantly improving the performance of variable stiffness actuators in complex environments, and is suitable for mechanical control systems with high-precision requirements, such as robotics, intelligent manufacturing and other fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 The present invention provides a characteristic diagram of a type A nonlinear gain function;

[0068] Figure 2 The present invention provides a characteristic diagram of a B-type nonlinear gain function;

[0069] Figure 3 The present invention provides a characteristic diagram of the stiffness control rate;

[0070] Figure 4 It is a schematic diagram of the setting process provided by the present invention;

[0071] Figure 5 This is a schematic diagram of nonlinear PID control parameter tuning provided by the present invention;

[0072] Figure 6 This is a schematic diagram of setting the stiffness control rate parameters provided by the present invention;

[0073] Figure 7 Schematic diagram of force control error simulation effect of force-stiffness control provided by the present invention;

[0074] Figure 8 Schematic diagram of the simulation of the speed change of the controlled system of the force-stiffness control provided by the present invention;

[0075] Figure 9 Schematic diagram of a force-stiffness closed-loop control system of a variable stiffness actuator based on force feedback provided by the present invention;

[0076] Figure 10 It is a system diagram of the method verification equipment provided by the present invention;

[0077] Figure 11 This is a surface accuracy diagram after the method provided by the present invention is applied to robot grinding and polishing;

[0078] Figure 12 This is a comparison chart of force control accuracy after the method provided by the present invention is applied to robot grinding and polishing. DETAILED DESCRIPTION

[0079] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0080] Example 1: Robot Arm Force-Stiffness Control System

[0081] In industrial robotics, especially in high-precision assembly and machining applications, robotic arms must precisely control the force and stiffness they apply to objects. Traditional robotic control systems typically achieve force control by fixing stiffness, but the accuracy and real-time responsiveness of stiffness control fall short in complex tasks, especially when handling fragile, flexible, or complex-shaped objects.

[0082] In this case study, the force-stiffness closed-loop control method for a variable-stiffness actuator based on force feedback, as proposed in this paper, was applied to the control system of a robotic arm. By integrating nonlinear PID control with closed-loop stiffness feedback, the system automatically adjusts force and stiffness based on task requirements, achieving real-time optimization during execution.

[0083] The specific steps include:

[0084] 1. The robot arm uses built-in sensors to measure the contact force and deformation of the object in real time.

[0085] 2. The robot arm's motion is regulated by a nonlinear PID control method, precisely controlling force and stiffness. This allows the robot to adjust force output based on the object's flexibility and shape changes, avoiding excessive force and damage to the object.

[0086] 3. The stiffness feedback control system combines real-time force feedback to dynamically adjust the stiffness of the actuator, ensuring that the robot can adapt to load changes under different operating conditions and ensuring the stability and accuracy of the operation process.

[0087] This embodiment greatly improves the adaptability and accuracy of the robot in complex operations by precisely controlling force and stiffness, effectively reduces the risk of damage during the production process, and improves production efficiency and product quality.

[0088] Example 2: Intelligent rehabilitation therapy equipment

[0089] In rehabilitation medicine, traditional rehabilitation equipment cannot automatically adjust force and stiffness based on individual patient differences and changes in condition, resulting in limited treatment effectiveness and difficulty in providing real-time feedback and adjustments. Especially for patients with spinal and joint injuries, force and stiffness during rehabilitation must be flexibly adjusted based on the patient's recovery progress to avoid overloading or applying inappropriate force.

[0090] The present invention's force-stiffness closed-loop control method for variable-stiffness actuators based on force feedback is applied to intelligent rehabilitation therapy equipment. By integrating nonlinear PID control and stiffness closed-loop control technology, the device achieves precise regulation and real-time feedback during treatment. The specific steps include:

[0091] 1. The device sets the initial force and stiffness based on the patient's recovery status and treatment plan, and uses sensors to monitor the patient's joint or spine responses, such as force and angle changes, in real time.

[0092] 2. The nonlinear PID control method adjusts the output force of the device according to the monitored error, and optimizes the proportional, integral and differential coefficients by adjusting the nonlinear gain function to ensure that the force output matches the patient's needs.

[0093] 3. The stiffness closed-loop control system uses real-time feedback to adjust the stiffness of the actuator to adapt to the patient's dynamic needs, ensuring the balance of force and stiffness during treatment and avoiding excessive stress or discomfort to the patient.

[0094] This embodiment provides personalized and precise rehabilitation treatment by precisely controlling the force and stiffness during the treatment process, thereby improving the patient's rehabilitation effect, reducing discomfort during the recovery process, and significantly improving treatment efficiency and effectiveness. It is particularly suitable for rehabilitation treatment of the spine, joints, etc.

[0095] An embodiment of the present invention provides a force-stiffness closed-loop control method and system for a variable stiffness actuator based on force feedback, including a nonlinear PID control method, a stiffness closed-loop control method, and a parameter tuning method and process.

[0096] The nonlinear PID control method and the stiffness closed-loop control method constitute a force-stiffness control system. The controlled system consists of a thrust generating mechanism and a variable stiffness mechanism, which jointly output force.

[0097] The nonlinear PID control method is obtained by constructing a nonlinear gain function and multiplying the nonlinear gain function with the linear PID gain coefficient, which is expressed as

[0098]

[0099] Where e is the error, which is the difference between the sensor measurement value and the expected value, K P Indicates the linear PID proportional coefficient, K I Indicates the linear PID integral coefficient, K D Indicates the linear PID differential coefficient, H P (e) represents the nonlinear gain function of the proportional coefficient and is related to the error, H I (e) represents the nonlinear gain function of the integral coefficient and is related to the error, H D (e) represents the nonlinear gain function of the differential coefficient and is related to the error, uc represents the output expectation of the controller, and t is time;

[0100] The nonlinear gain function includes two forms, A and B, both of which are constructed by hyperbolic tangent function;

[0101] The A-type nonlinear gain function expression is H A (e) = 1 + h 11 (1-sech(h 12 e)), H A (e) is the value of the nonlinear gain function; h 11 and h 12 is the nonlinear parameter of the type I nonlinear gain function; e is the error, which is expressed as the difference between the sensor measurement value and the expected value;

[0102] like Figure 1 As shown, the A-type nonlinear gain function tends to 1 when the feedback error e tends to 0; when the feedback error e tends to infinity, the nonlinear gain function tends to 1+h 11 ;

[0103] The B-type nonlinear gain function is H B (e) is the value of the nonlinear gain function; h 21 and h 22is the nonlinear parameter of the B-type nonlinear gain function; e is the error, which is expressed as the difference between the sensor measurement value and the expected value;

[0104] like Figure 2 As shown, the B-type nonlinear gain function, when the feedback error e tends to 0, the nonlinear gain function tends to 1; when the feedback error e tends to infinity, the nonlinear gain function tends to

[0105] By combining type A and type B nonlinear gain functions, a variety of nonlinear gain functions can be obtained; the nonlinear gain functions in the order of proportion, integration, and differentiation can be expressed as nine combinations: AAA, AAB, ABA, ABB, BAA, BAB, ABB, BBA, and BBB;

[0106] A typical example of nonlinear PID combination can be expressed as:

[0107]

[0108] The stiffness closed-loop control method is based on a closed-loop stiffness feedback rate constructed by force feedback, tangent function, and exponential function, and changes the system stiffness by controlling the variable stiffness module through force feedback;

[0109] The closed-loop stiffness feedback rate discrete time domain expression:

[0110]

[0111] Among them, k vs (t) represents the expected stiffness at time t, k vs (t-1) represents the desired stiffness at time t, T is the discrete period of the control system, Δe represents the forward difference of the error e at time t, α1, α2, and α3 are the adjustment parameters of the stiffness control rate, and e is the error, which is expressed as the difference between the sensor measurement value and the expected value.

[0112] like Figure 3 As shown, the closed-loop stiffness feedback rate is based on Evaluate the current rate of change of the force error; use the hyperbolic tangent function instead of the sign function to reduce chatter in stiffness control caused by force error noise interference;

[0113] If the force error change rate is positive, the stiffness is increased to reduce the damping ratio and response time, thereby enhancing the system's ability to quickly adjust to the force error. If the force error change rate is negative, the stiffness is reduced to increase the damping ratio and improve system stability, thereby enhancing the system's anti-interference performance.

[0114] like Figure 3 As shown, the magnitude of the stiffness change depends on the exponential decay function (1-exp(α3|e f(t)|)); As the absolute error changes, larger force errors lead to larger adjustments, while smaller errors produce smoother changes.

[0115] like Figure 4 As shown, the parameter setting method and process are as follows:

[0116] S1 is identified by the control system

[0117] Linear parameter identification of S2 nonlinear PID: K P , K I , K D

[0118] Nonlinear gain function parameter identification of S3 nonlinear PID: H P (e), H I (e), H D (e)

[0119] S4 stiffness control law parameter identification: α1, α2 and α3

[0120] Furthermore, the parameters of the controlled system in S1 can be identified by using the sweep frequency method and the pseudo-random sequence method to obtain the transfer function of the system;

[0121] The linear parameter identification in S2 can be done by empirical accumulation method, Ziegler-Nichols method, and Cohen-Coon method;

[0122] like Figure 5 As shown in Figure 2, the nonlinear gain function parameter identification in S3 is implemented using the particle swarm optimization algorithm. Other algorithms such as genetic algorithms can also be used, but the algorithm optimization goal is limited to the weighted sum of the output of the controlled system and the variance of the nonlinear PID control output as the objective function, which is expressed as follows:

[0123] Where α and β are weights, and J is the optimization target; The variance of the output of the controlled system is used to judge the stability of the system output; The variance of the nonlinear PID control output is used to evaluate the anti-disturbance capability of the nonlinear PID control system. When the controlled system is relatively stable and the external disturbance is relatively small, a larger α value can be set. When the external disturbance is relatively large, a larger β value can be set.

[0124] like Figure 5 As shown in Figure 3, during the parameter tuning process of the heuristic optimization algorithm in S3, the expected value of the controlled system output is set to 0, a random disturbance is set in the output of the controlled system, and the stiffness control law is not applied; the parameter group corresponding to the minimum value of the objective function is selected as the parameter tuning result.

[0125] like Figure 6As shown in Figure 2, the stiffness control law parameter identification in S4 is implemented using a heuristic optimization algorithm, such as a particle swarm optimization algorithm or a genetic algorithm. However, the algorithm optimization goal is limited to the weighted sum of the output of the controlled system and the variance of the control system output as the objective function, which is expressed as follows:

[0126] Where α and β are weights, and J is the optimization target; The variance of the output of the controlled system is used to judge the stability of the system output; is the variance of the nonlinear PID control output, The stiffness control rate controls the output variance and is used to evaluate the anti-disturbance capability of the entire control system;

[0127] like Figure 6 As shown, in the parameter tuning process of the heuristic optimization algorithm in S4, the expected value of the controlled system output is set to 0, and a random disturbance is set in the output of the controlled system; the parameter group corresponding to the minimum value of the objective function is selected as the tuning parameter result.

[0128] like Figure 7 As shown in the figure, the force control error of the method proposed in this paper is reduced in a strong disturbance environment. A force-stiffness closed-loop control method and system of a variable stiffness actuator based on force feedback can effectively improve the force control accuracy, and as shown in the figure, Figure 8 As shown in Figure 2, the speed fluctuation of the actuator (controlled system) has dropped dramatically, which means that the control stability of the system has become better.

[0129] like Figure 9 As shown, an embodiment of the present invention provides a force-stiffness closed-loop control system for a variable stiffness actuator based on force feedback based on the force-stiffness closed-loop control method for a variable stiffness actuator based on force feedback, the system comprising:

[0130] The nonlinear PID module is obtained by constructing a nonlinear gain function and multiplying the nonlinear gain function with the linear PID gain coefficient;

[0131] The stiffness control rate module is based on the closed-loop stiffness feedback rate constructed by force feedback, tangent function, and exponential function. It controls the variable stiffness module through force feedback to change the system stiffness.

[0132] Parameter tuning module, identified by the control system; linear parameter identification of nonlinear PID: K P , K I , K D ; Nonlinear gain function parameter identification of nonlinear PID: H P (e), H I (e), H D (e); Identification of stiffness control law parameters: α1, α2 and α3.

[0133] An embodiment of the present invention provides a computer device, which includes a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor executes the steps of the force-stiffness closed-loop control method of a variable stiffness actuator based on force feedback.

[0134] An embodiment of the present invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the force-stiffness closed-loop control method of a variable stiffness actuator based on force feedback.

[0135] An embodiment of the present invention provides an information data processing terminal, which is used to implement the force-stiffness closed-loop control system of the variable stiffness actuator based on force feedback.

[0136] This invention has been successfully verified on a robotic grinding and polishing experimental platform. Experiments on curved, thin-walled parts have been conducted, and the equipment system is shown in the figure below.

[0137] Figure 10 It is a system diagram of the method verification equipment provided by the present invention.

[0138] The following figure shows the experimental results provided by the present invention. Compared with PID control, the average absolute force error, maximum absolute force error, material removal error and surface roughness of the method of the present invention are reduced by 69.78%, 64.6%, 70.8% and 65.49%, respectively.

[0139] Figure 11 This is a surface accuracy diagram after the method provided by the present invention is applied to robot grinding and polishing.

[0140] Figure 12 This is a comparison chart of force control accuracy after the method provided by the present invention is applied to robot grinding and polishing.

[0141] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware portion can be implemented using dedicated logic; the software portion can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art will appreciate that the above-mentioned devices and methods can be implemented using computer-executable instructions and / or contained in processor control code, for example, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, or can be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0142] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with this technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered by the scope of protection of the present invention.

Claims

1. A force-stiffness closed-loop control method for a variable stiffness actuator based on force feedback, characterized in that: The method includes a nonlinear PID control method, a stiffness closed-loop control method, and a parameter tuning method and process; The parameter setting method and process are as follows: S1: Identified by the control system S2: Linear parameter identification of nonlinear PID: K P , K I , K D S3: Nonlinear gain function parameter identification of nonlinear PID: H P (e), H I (e), H D (e) S4: Identification of stiffness control law parameters: α1, α2, and α3 The parameters of the controlled system in S1 can be identified by using the sweep frequency method or the pseudo-random sequence method to obtain the transfer function of the system; The linear parameter identification in S2 can be carried out by the empirical accumulation method, the Ziegler-Nichols method, and the Cohen-Coon method; The nonlinear gain function parameter identification in S3 is implemented by a heuristic optimization algorithm, but the algorithm optimization target is limited to the weighted sum of the output of the controlled system and the variance of the nonlinear PID control output as the objective function, which is expressed as follows: Where α and β are weights, and J is the optimization target; The variance of the output of the controlled system is used to judge the stability of the system output; The variance of the nonlinear PID control output is used to evaluate the anti-disturbance capability of the nonlinear PID control system. When the controlled system is relatively stable and the external disturbance is relatively small, a larger α value can be set. When the external disturbance is relatively large, a larger β value can be set. During the parameter tuning process of the heuristic optimization algorithm in S3, the output expected value of the controlled system is set to 0, a random disturbance is set in the output of the controlled system, and the stiffness control law is not applied; The stiffness control law parameter identification in S4 is realized by using a heuristic optimization algorithm, but the algorithm optimization goal is limited to the weighted sum of the output of the controlled system and the variance of the control system output as the objective function, which is expressed as follows: Where α and β are weights, and J is the optimization target; The variance of the output of the controlled system is used to judge the stability of the system output; is the variance of the nonlinear PID control output, The stiffness control rate controls the output variance and is used to evaluate the anti-disturbance capability of the entire control system; During the parameter tuning process of the heuristic optimization algorithm in S4, the output expected value of the controlled system is set to 0, and a random disturbance is set in the output of the controlled system.

2. The force-stiffness closed-loop control method for a variable stiffness actuator based on force feedback according to claim 1, characterized in that: The nonlinear PID control method is obtained by constructing a nonlinear gain function and multiplying the nonlinear gain function with the linear PID gain coefficient, which is expressed as Where e is the error, which is the difference between the sensor measurement value and the expected value, K P Indicates the linear PID proportional coefficient, K I Indicates the linear PID integral coefficient, K D Indicates the linear PID differential coefficient, H P (e) represents the nonlinear gain function of the proportional coefficient and is related to the error, H I (e) represents the nonlinear gain function of the integral coefficient and is related to the error, H D (e) represents the nonlinear gain function of the differential coefficient and is related to the error, u c represents the output expectation of the controller, and t is the time.

3. The force-stiffness closed-loop control method for a variable stiffness actuator based on force feedback as claimed in claim 2, characterized in that: The nonlinear gain function includes two forms, A and B, both of which are constructed by hyperbolic tangent function; The expression of the A-type nonlinear gain function is H A (e) = 1 + h 11 (1-sech(h 12 e)), H A (e) is the value of the nonlinear gain function; h 11 and h 12 is the nonlinear parameter of the type I nonlinear gain function; e is the error, which is expressed as the difference between the sensor measurement value and the expected value; Type A nonlinear gain function, when the feedback error e tends to 0, the nonlinear gain function tends to 1; When the feedback error e tends to infinity, the nonlinear gain function tends to 1+h 11 ; The B-type nonlinear gain function is is the value of the nonlinear gain function; h 21 and h 22 is the nonlinear parameter of the B-type nonlinear gain function; e is the error, which is expressed as the difference between the sensor measurement value and the expected value; Type B nonlinear gain function, when the feedback error e tends to 0, the nonlinear gain function tends to 1; When the feedback error e tends to infinity, the nonlinear gain function tends to By combining type A and type B nonlinear gain functions, a variety of nonlinear gain functions can be obtained; the nonlinear gain functions in the order of proportion, integration, and differentiation can be expressed as eight combinations: AAA, AAB, ABA, ABB, BAA, BAB, BBA, and BBB.

4. The force-stiffness closed-loop control method for a variable stiffness actuator based on force feedback according to claim 1, characterized in that: The stiffness closed-loop control method is based on a closed-loop stiffness feedback rate constructed by force feedback, tangent function, and exponential function, and changes the system stiffness by controlling the variable stiffness module through force feedback; The closed-loop stiffness feedback rate discrete time domain expression: Among them, k vs (t) represents the expected stiffness at time t, k vs (t-1) represents the desired stiffness at time t, T is the discrete period of the control system, Δe represents the forward difference of the error e at time t, α1, α2, and α3 are the adjustment parameters of the stiffness control rate, and e is the error, which is expressed as the difference between the sensor measurement value and the expected value; The closed-loop stiffness feedback rate is based on Evaluate the current rate of change of the force error; use a hyperbolic tangent function instead of a sign function to reduce chatter in stiffness control caused by force error noise interference; if the force error rate of change is positive, increase the stiffness to reduce the damping ratio and response time, thereby enhancing the system's ability to quickly adjust to the force error; if the force error rate of change is negative, decrease the stiffness to increase the damping ratio and improve system stability, thereby enhancing the system's anti-interference performance; The magnitude of the stiffness change depends on the exponential decay function (1-exp(α3|e f (t)|)); As the force error changes, larger force errors lead to larger adjustments, while smaller errors produce smoother changes.

5. A force-stiffness closed-loop control system for a variable stiffness actuator based on force feedback, based on the force-stiffness closed-loop control method for a variable stiffness actuator based on force feedback as claimed in any one of claims 1 to 4, characterized in that: The system includes: The nonlinear PID module is obtained by constructing a nonlinear gain function and multiplying the nonlinear gain function with the linear PID gain coefficient; The stiffness control rate module is based on the closed-loop stiffness feedback rate constructed by force feedback, tangent function, and exponential function. It controls the variable stiffness module through force feedback to change the system stiffness. Parameter tuning module, identified by the control system; linear parameter identification of nonlinear PID: K P , K I , K D ; Nonlinear gain function parameter identification of nonlinear PID: H P (e), H I (e), H D (e); Identification of stiffness control law parameters: α1, α2 and α3.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the force-stiffness closed-loop control method of a variable stiffness actuator based on force feedback as described in any one of claims 1 to 4.

7. A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the force-stiffness closed-loop control method of a variable stiffness actuator based on force feedback as described in any one of claims 1 to 4.

8. An information data processing terminal, characterized in that: The information data processing terminal is used to implement the variable stiffness actuator force-stiffness closed-loop control system based on force feedback as described in claim 5.

Images