Method for intelligent control of preset performance of electromechanical position servo system

Abstract

The application discloses a kind of electromechanical position servo system preset performance intelligent control method, this control method fuses the thought of backstepping control, designs the preset performance nonlinear position controller of combination nonlinear filtering, neural network adaptation, disturbance suppression.For the position control problem of electromechanical servo system, the preset time convergence of tracking error is realized by designing preset performance function, the transient and steady-state performance of system is guaranteed, and based on neural network adaptation, nonlinear filter is constructed Robust controller, effectively handle the nonlinear characteristics and model uncertainty existing in system, high-precision asymptotic tracking performance of system is realized.The method proposed does not need to assume that unknown disturbance is continuous and derivable, effectively reduces the complexity of traditional robust controller design.

Description

Technical Field

[0001] This invention relates to the field of electro-hydraulic servo control technology, specifically to an intelligent control method for preset performance of electromechanical position servo systems (RBFPPC) based on barrier Lyapunov functions, nonlinear dynamic surface control filters, and radial basis function neural network adaptive laws. Background Technology

[0002] Electromechanical position servo systems, characterized by their fast response, high integration, and high reliability, are widely used in inertial navigation testing, radar tracking, and optical aiming. An electromechanical position servo system can be described as an electromechanical system composed of a control system, drive motor, transmission mechanism, and load; it is a typical strongly nonlinear system. The system exhibits uncertainties in parameters such as friction and coupling forces, as well as uncertain nonlinearities such as modeling errors that are difficult to characterize precisely, unmodeled external disturbances, and nonlinear friction, which severely restrict the system's control performance. With increasing industrialization, electromechanical position servo systems are being used more and more extensively in high-precision servo systems. To meet the complex and diverse tasks and high-performance requirements, higher demands are being placed on performance indicators such as transient and steady-state tracking accuracy. Traditional model-free linear control, such as PID control, while widely used, struggles to meet the demands of high-precision tracking performance. Therefore, researching advanced nonlinear control based on system dynamic models is of great significance.

[0003] Numerous research achievements have been made on the nonlinear control problems of servo systems such as electromechanical position servo systems. Among them, model-based control methods such as feedback linearization control, adaptive backstepping control, and adaptive robust control have shown good robustness and can provide effective solutions for high-precision trajectory tracking control. However, due to the large amount of uncertainty in the model, their control performance is greatly limited. Adaptive control can effectively handle most uncertainties, but its ability to handle unmodeled disturbances is poor. Traditional adaptive robust control can effectively deal with the high nonlinearity of the system and handle most model uncertainties and uncertain nonlinearities, but the uncertain disturbances in the system may lead to potentially high-gain feedback designs, and the increase in system order can easily cause the differential explosion problem of controllers designed based on the backstepping method. Adaptive control based on error sign integral robustness can reduce the feedback burden of the controller and obtain asymptotic tracking performance; however, this controller cannot handle mismatched disturbances and cannot completely solve the control problems of turntable systems. Active disturbance rejection control can realize model compensation for matched and mismatched disturbances in high-order systems and improve the robustness of the system, but it is difficult to achieve asymptotic tracking performance within a finite time. Furthermore, most of the above nonlinear controllers pay little attention to the convergence speed performance of tracking errors, making it difficult to obtain the preset transient and steady-state performance, and still have significant limitations in practical applications. Summary of the Invention

[0004] The purpose of this invention is to provide a position control method for an electromechanical position servo system that has neural network adaptive compensation capability, error transient and steady-state performance preset capability, and asymptotic tracking performance. It can not only guarantee the transient and steady-state performance of tracking error by designing preset performance functions, but also construct a controller based on radial basis neural network and nonlinear filtering to effectively handle the unknown nonlinearity in the system and realize high-precision asymptotic tracking control of the system.

[0005] The technical solution to achieve the objective of this invention is: a method for intelligent control of preset performance of an electromechanical position servo system, comprising the following steps:

[0006] Step 1: Establish the mathematical model of the electromechanical position servo system, then proceed to Step 2.

[0007] Step 2: Based on the mathematical model of the electromechanical position servo system, design a pre-set performance intelligent controller, and proceed to Step 3.

[0008] Step 3: Apply Lyapunov stability theory to perform stability analysis on the designed pre-set performance intelligent controller, and obtain the result that the system tracking error is asymptotically stable.

[0009] Compared with the prior art, the significant advantages of this invention are: (1) nonlinear feedforward compensation is constructed by building a model based on radial basis function neural network; (2) the transient and steady-state performance of the system is controlled by designing a preset performance function for the error and constructing a barrier Lyapunov function; (3) the introduction of a nonlinear dynamic surface control filter avoids the differentiation of the virtual control input, solves the differential explosion problem in traditional backstep control, and improves the practicality of the controller; (4) the designed nonlinear robust term realizes smooth continuous control through a continuous tanh function, improves the practicality of the designed controller in engineering practice, and the simulation results verify its effectiveness; (5) there is no need to assume that the unknown disturbance is continuous and differentiable, which effectively reduces the complexity of traditional robust controller design. Attached Figure Description

[0010] Figure 1 This is a schematic diagram illustrating the principle of the intelligent control method for preset performance of the electromechanical position servo system of the present invention.

[0011] Figure 2 This is a simplified schematic diagram of the electromechanical position servo system of the present invention.

[0012] Figure 3 This is a position tracking curve diagram under the action of the RBFPPC controller designed in this invention.

[0013] Figure 4This is a comparison curve of the tracking error under the action of the preset performance intelligent controller RBFPPC designed in this invention and the traditional PID method.

[0014] Figure 5 This is a speed tracking error curve under the action of the preset performance intelligent controller RBFPPC designed in this invention.

[0015] Figure 6 This is a curve showing the upper bound estimation of the disturbance under the action of the RBFPPC controller designed in this invention.

[0016] Figure 7 This is a system control input curve diagram under the action of the RBFPPC controller designed in this invention. Detailed Implementation

[0017] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] Combination Figure 1 and Figure 2 The intelligent control method for preset performance of the electromechanical position servo system described in this invention includes the following steps:

[0019] Step 1: Establish a mathematical model of the electromechanical position servo system.

[0020] Step 1-1: The electromechanical position servo system is applied to electromechanical servo equipment that requires high-precision angular position tracking in scenarios such as inertial navigation testing, radar tracking, and optical aiming. Taking a turntable as an example, the turntable load is fixedly connected to the output shaft of the drive motor, and the drive motor drives the load to rotate. Based on the dynamic characteristics of the turntable load, drive motor, and transmission mechanism, a mathematical model of the electromechanical position servo system is established.

[0021] According to Newton's equations of motion, the torque balance equation of the electromechanical position servo system is expressed as:

[0022] (1),

[0023] In equation (1), This represents the lumped moment of inertia of the turntable system about its axis of rotation. This represents the angular acceleration of the turntable. This represents the electromagnetic torque output by the drive motor. This represents the frictional torque acting on the turntable system. This represents the torque caused by an unknown external disturbance to the system. Indicates time.

[0024] Frictional torque on the turntable system Specifically defined as:

[0025] (2),

[0026] In equation (2), Indicates the coefficient of viscous friction. This represents the modelable Coulomb friction amplitude. This indicates the additional amplitude of Stribeck's static friction. This represents Stribeck's critical velocity. Represents the coefficients of the tanh curve. This indicates the angular velocity of the turntable. Represents the natural constant.

[0027] Ignoring the electrical dynamics of the drive motor and the elastic deformation of the transmission mechanism, define the system control input. For the voltage signal applied to the drive motor, then With the electromagnetic torque output by the drive motor A proportional relationship, that is, satisfying:

[0028] (3),

[0029] In the formula, This indicates the torque coefficient of the drive motor.

[0030] Substituting equations (2) and (3) into equation (1), we obtain the comprehensive physical model of the electromechanical position servo system:

[0031] (4).

[0032] Steps 1-2: Define the state variables required for controller design and reconstruct the mathematical model of the electromechanical position servo system into state-space equation form.

[0033] Define state variables: ,in, Indicates the corner position of the turntable.

[0034] Combining equations (1) to (4), the system state-space equations are rewritten in the following form:

[0035] (5),

[0036] In equation (5), express The derivative, express The derivative, This represents the control gain that is known to be always positive. , The unknown nonlinear function representing the integration of viscous friction, Coulomb friction, and Stribeck friction is given. , This indicates an unknown external disturbance. .

[0037] The following assumptions are made for the electromechanical position servo system:

[0038] Assumption 1: The system expects to track a trajectory The second derivative is bounded.

[0039] Assumption 2: Unknown external disturbances It is bounded, satisfying ,in It represents an unknown nonnegative constant.

[0040] Assumption 3: Unknown nonlinear function In close proximity The above can be approximated by a radial basis function neural network.

[0041] Assumption 4: There exists a positive integrable function. and The following conditions must be met:

[0042] (6),

[0043] In equation (6), Represents positive numbers. Represents positive numbers. This represents the integral variable.

[0044] Proceed to step 2.

[0045] Step 2: Based on the mathematical model of the electromechanical position servo system, design a pre-defined performance intelligent controller.

[0046] Step 2-1: To solve the unknown nonlinear problems existing in the system, an adaptive law of radial basis function neural network is constructed to achieve the adaptation of unknown nonlinear functions. Accurate estimation of the RBF neural network feedforward compensation term was designed.

[0047] Based on assumption 3, the unknown nonlinear function It can be approximated by a radial basis function neural network. Reconstructed into the following form:

[0048] (7),

[0049] In equation (7), Represents the ideal neural network weight vector, with superscript... This indicates the transpose operation. Let the known radial basis function vector be the first... The output of the radial basis functions Represented as ,in , Indicates the number of hidden layer nodes. Indicates the first The center of each basis function This represents the width of the basis functions. This represents the natural exponential function. The neural network approximation error is expressed as follows: ,in This represents a nonnegative bounded constant.

[0050] definition Represents the weights of an ideal neural network The estimated value is used to construct the feedforward compensation term of the RBF neural network. To achieve the understanding of unknown nonlinear functions Feedforward compensation.

[0051] The adaptive weight law of the RBF neural network is designed in the following form:

[0052] (8),

[0053] In equation (8), express The first derivative, This represents the positive definite adaptive gain matrix. This represents the weighted adaptive function.

[0054] Step 2-2: To ensure both steady-state and transient performance of the tracking, and to ensure that the tracking error converges rapidly to a specified range within a specified time, define the tracking error. and Design a preset performance function Construct error preset boundaries and construct transformation error. To replace tracking error By guaranteeing The convergence performance thus ensures Its convergence performance.

[0055] To achieve preset performance control of tracking error, position tracking error is defined. Velocity layer error ,in Indicates virtual control The filtered signal.

[0056] Design the following preset performance function. :

[0057] (9),

[0058] In equation (9), All are positive adjustable parameters and satisfy... , Indicates position tracking error and speed tracking error initial value, This represents the initial value of the preset performance function. This indicates the maximum permissible steady-state error. This indicates the preset convergence rate.

[0059] Define error transformation variables :

[0060] (10),

[0061] Its derivative Represented as:

[0062] (11),

[0063] In equation (11), Indicates error The first derivative, Indicates the preset performance function The first derivative.

[0064] Steps 2-3: Integrating the concept of backstepping control, a model feedforward term is constructed based on the designed RBF neural network. The predetermined convergence characteristics of the position tracking error are guaranteed by the designed preset performance function. A barrier Lyapunov function is constructed as an analysis tool, and the convergence performance of each error stage is further guaranteed by introducing a nonlinear robust term. Based on this, virtual control laws are designed stage by stage. and actual control law The filtered signal for the virtual control input is designed using a nonlinear dynamic surface control filter. This enables control over the system.

[0065] According to the state-space equation, the position tracking error derivative Represented as:

[0066] (12).

[0067] To avoid the differential explosion problem in the system, a virtual control input is designed for the controller. Design the following nonlinear dynamic surface control filter:

[0068] (13),

[0069] In equation (13), Indicates the filtered signal The first derivative, This represents the positive damping parameter of the adjustable filter. Indicates the filtering error. Let represent a positive integrable function. express The upper bound estimate, of which express The first derivative, Updated by the following formula:

[0070] (14),

[0071] In equation (14), express The derivative, This indicates a positive adaptive gain.

[0072] Constructing barrier Lyapunov functions :

[0073] (15),

[0074] right Differentiation yields :

[0075] (16),

[0076] exist Then, substitute equations (11) and (12) into equation (16):

[0077] (17),

[0078] Based on this, a virtual control law is designed. for:

[0079] (18),

[0080] In equation (18), This represents the model's feedforward compensation term. Represents the linear robust feedback term. This represents a positive feedback gain.

[0081] According to the state-space equation, the velocity tracking error derivative Represented as:

[0082] (19),

[0083] exist Then, substitute equations (11) and (19) into equation (16):

[0084] (20).

[0085] Based on this, the actual control law is designed. for:

[0086] (twenty one),

[0087] In equation (21), This represents the model's feedforward compensation term. Represents the linear robust feedback term. Represents the nonlinear robust term. This represents the feedforward compensation term in the RBF neural network. This represents a positive feedback gain.

[0088] Design nonlinear robust terms for:

[0089] (twenty two),

[0090] In the formula, Describing the aggregate bounded disturbance upper bound value The estimate, of which , , It represents a positive integrable function.

[0091] Updated by the following formula:

[0092] (twenty three),

[0093] In the formula, express The first derivative, This indicates a positive adaptive gain.

[0094] Proceed to step 3.

[0095] Step 3: Apply Lyapunov stability theory to perform stability analysis on the designed pre-set performance intelligent controller and obtain the result that the system tracking error is asymptotically stable.

[0096] Constructing Lyapunov functions as follows:

[0097] (twenty four),

[0098] In equation (24), Both represent intermediate variables. The upper bound estimation error of the lumped disturbance is expressed as... , The first derivative of the virtual control law The upper bound of the estimation error, written as , The error in RBF weight estimation is written as: .

[0099] Combined equations (17) and (18), time derivative for:

[0100] (25).

[0101] Combined equations (20) and (21), time derivative for:

[0102] (26).

[0103] Combined equations (8), (14), (22), (23), (25), and (26). time derivative Represented as:

[0104] (27),

[0105] Use the following inequalities to... Scaling process:

[0106] (28),

[0107] In equation (28), Both represent constants.

[0108] By combining equations (27) and (28), we obtain Inequalities:

[0109] (29),

[0110] Furthermore, equation (29) can be reconstructed as follows:

[0111] (30),

[0112] In equation (30), the error vector , The positive definite gain matrix can be represented as:

[0113] (31),

[0114] Furthermore, equation (30) can be written as:

[0115] (32),

[0116] In equation (32), Representation matrix The smallest eigenvalue, This represents a positive function.

[0117] Integrating both sides of equation (32) with respect to time, according to assumption 4, we get:

[0118] (33),

[0119] From equation (33), we can see that and ,then Both are bounded, and according to the properties of the preset performance function, the state error can be obtained. Both are bounded. Based on Assumption 1 and Assumption 2, the system state is... Both are bounded. According to Assumption 4, the control input is bounded. Since it is bounded, all closed-loop signals are bounded.

[0120] At the same time, according to the error The dynamics can be obtained The boundedness of therefore we know The uniform continuity. Based on Barbarat's lemma, it can be known that when hour, Therefore, the following conclusion can be drawn:

[0121] The proposed control law can achieve asymptotic output tracking, that is, when , This ensures the boundedness of all closed-loop system signals. A schematic diagram of the system's preset performance control method is shown below. Figure 1 As shown.

[0122] Example

[0123] To evaluate the performance of the designed controller, the physical parameters of the electromechanical position servo system in the simulation are shown in Table 1:

[0124] Table 1 System Physical Parameters

[0125]

[0126] Given the desired instructions of the system The initial position of the system is The initial velocity is .

[0127] The following controller is used for comparison in the simulation:

[0128] Preset Performance Intelligent Controller (RBFPPC): Take controller gain , ; Take the number of hidden layer nodes in the neural network basis function width Weighted adaptive gain ; Retrieve preset performance function parameters , , , , , Take the parameters of the nonlinear filter. ; Adaptive gain , Take a positive integrable function , .

[0129] PID controller: The selected controller parameters are as follows , , .

[0130] The curves showing the desired and actual system trajectory tracked by the RBFPPC controller, the comparison curves of position tracking errors under the action of the RBFPPC controller and the PID controller, the speed tracking error curve under the action of the RBFPPC controller, the disturbance upper bound estimation curve under the action of the RBFPPC controller, and the system control input curve under the action of the RBFPPC controller are respectively as follows: Figure 3 , Figure 4 , Figure 5 , Figure 6 , Figure 7 As shown. By Figure 4 It can be seen that the electromechanical position servo system achieved good tracking performance under the action of the RBFPPC controller, and the amplitude of its steady-state tracking error absolute value is approximately The mean is approximately Furthermore, it can be quickly and effectively constrained by preset performance boundaries, achieving rapid convergence. Its tracking performance is far superior to PID control, and PID control cannot guarantee that the tracking error will always converge to within the preset boundaries. Figure 6 It can be seen that the estimated value of the upper bound of the perturbation gradually converges under the action of the adaptive law. Figure 7 It can be seen that the designed controller has a smoother and more continuous control input, which is more conducive to the controller's execution in practical applications.

Claims

1. A method for intelligent control of preset performance in an electromechanical position servo system, characterized in that, Includes the following steps: Step 1: Establish a mathematical model of the electromechanical position servo system. ， in, This represents the lumped moment of inertia of the turntable system about its axis of rotation. This represents the angular acceleration of the turntable. This indicates the torque coefficient of the drive motor. For system control input, Indicates the coefficient of viscous friction. This indicates the angular velocity of the turntable. This represents the modelable Coulomb friction amplitude. Represents the coefficients of the tanh curve. This indicates the additional amplitude of Stribeck's static friction. This represents Stribeck's critical velocity. This represents the torque caused by an unknown external disturbance to the system. Indicate the time, proceed to step 2; Step 2: Based on the mathematical model of the electromechanical position servo system, design a pre-defined performance intelligent controller, and proceed to Step 3; Step 3: Apply Lyapunov stability theory to perform stability analysis on the designed pre-set performance intelligent controller, and obtain the result that the system tracking error is asymptotically stable.

2. The intelligent control method for preset performance of the electromechanical position servo system according to claim 1, characterized in that, In step 1, the mathematical model of the electromechanical position servo system is established, as follows: Step 1-1: The electromechanical position servo system is applied to electromechanical servo equipment that requires high-precision position tracking in inertial navigation testing, radar tracking, and optical aiming scenarios. The turntable load is fixedly connected to the output shaft of the drive motor, and the drive motor drives the load to rotate. Based on the dynamic characteristics of the turntable load, drive motor, and transmission mechanism, a mathematical model of the electromechanical position servo system is established. Steps 1-2: Define the state variables required for controller design and reconstruct the mathematical model of the electromechanical position servo system into state-space equation form.

3. The intelligent control method for preset performance of the electromechanical position servo system according to claim 2, characterized in that, In step 1-1, the details are as follows: According to Newton's equations of motion, the torque balance equation of an electromechanical position servo system is defined as follows: (1)， In equation (1), This represents the lumped moment of inertia of the turntable system about its axis of rotation. This represents the angular acceleration of the turntable. This represents the electromagnetic torque output by the drive motor. This represents the frictional torque acting on the turntable system. This represents the torque caused by an unknown external disturbance to the system. Indicates time; Frictional torque on the turntable system Specifically defined as: (2)， In equation (2), Indicates the coefficient of viscous friction. This represents the modelable Coulomb friction amplitude. This indicates the additional amplitude of Stribeck's static friction. This represents Stribeck's critical velocity. Represents the coefficients of the tanh curve. This indicates the angular velocity of the turntable. Represents the natural constant; Ignoring the electrical dynamics of the drive motor and the elastic deformation of the transmission mechanism, define the system control input. For the voltage signal applied to the drive motor, then With the electromagnetic torque output by the drive motor A proportional relationship, that is, satisfying: (3)， In the formula, This indicates the torque coefficient of the drive motor; Substituting equations (2) and (3) into equation (1), we obtain the comprehensive physical model of the electromechanical position servo system: (4)。 4. The intelligent control method for preset performance of the electromechanical position servo system according to claim 3, characterized in that, Steps 1-2 are detailed below: Define state variables : ,in, Indicates the corner position of the turntable; By combining equations (1) to (4), the system state-space equations are reconstructed into the following form: (5)， In equation (5), express The derivative, express The derivative, This represents the control gain that is known to be always positive. , The unknown nonlinear function representing the integration of viscous friction, Coulomb friction, and Stribeck friction is given. , This indicates an unknown external disturbance. .

5. The intelligent control method for preset performance of the electromechanical position servo system according to claim 4, characterized in that, In step 1, the following assumptions are made for the electromechanical position servo system: Assumption 1: The system expects to track a trajectory The second derivative is bounded; Assumption 2: Unknown external disturbances It is bounded, satisfying ,in Represents an unknown nonnegative constant; Assumption 3: Unknown nonlinear function In close proximity The above can be approximated by a radial basis function neural network; Assumption 4: There exists a positive integrable function. and The following conditions must be met: (6)， In equation (6), Represents positive numbers. Represents positive numbers. This represents the integral variable.

6. The intelligent control method for preset performance of the electromechanical position servo system according to claim 5, characterized in that, In step 2, based on the mathematical model of the electromechanical position servo system, a pre-defined performance intelligent controller is designed, as follows: Step 2-1: To solve the unknown nonlinear problems existing in the system, an adaptive law of radial basis function neural network is constructed to achieve the adaptation of unknown nonlinear functions. Accurate estimation of the RBF neural network feedforward compensation term was designed. Step 2-2: To ensure both steady-state and transient performance of the tracking, and to ensure that the tracking error converges rapidly to a specified range within a specified time, define the tracking error. and Design a preset performance function Construct error preset boundaries and construct transformation error. To replace tracking error By guaranteeing The convergence performance thus ensures Convergence performance; Steps 2-3: Integrating the concept of backstepping control, a model feedforward term is constructed based on the designed RBF neural network. The predetermined convergence characteristics of the position tracking error are guaranteed by the designed preset performance function. A barrier Lyapunov function is constructed as an analysis tool, and the convergence performance of each error stage is further guaranteed by introducing a nonlinear robust term. Based on this, virtual control laws are designed stage by stage. and actual control law The filtered signal for the virtual control input is designed using a nonlinear dynamic surface control filter. This enables control over the system.

7. The intelligent control method for preset performance of the electromechanical position servo system according to claim 6, characterized in that, In step 2-1, the specific details are as follows: Based on assumption 3, the unknown nonlinear function It can be approximated by a radial basis function neural network. Reconstructed into the following form: (7)， In equation (7), Represents the ideal neural network weight vector, with superscript... This indicates the transpose operation. Let the known radial basis function vector be the first... The output of the radial basis functions Represented as ,in , Indicates the number of hidden layer nodes. Indicates the first The center of each basis function This represents the width of the basis functions. Represents state variables, This represents the natural exponential function. The neural network approximation error is expressed as follows: , Represents a nonnegative bounded constant; definition For ideal neural network weights The estimated value is used to construct the feedforward compensation term of the RBF neural network. To achieve the understanding of unknown nonlinear functions Feedforward compensation; The adaptive weight law of the RBF neural network is designed in the following form: (8)， In equation (8), express The first derivative, This represents the positive definite adaptive gain matrix. This represents the weighted adaptive function.

8. The intelligent control method for preset performance of the electromechanical position servo system according to claim 7, characterized in that, In step 2-2, the specific details are as follows: To achieve preset performance control of tracking error, position tracking error is defined. Velocity layer error ,in Representing virtual control law Filtered signal; Design the following preset performance function. : (9)， In equation (9), All are positive adjustable parameters and satisfy... , Indicates position tracking error and speed tracking error initial value, This represents the initial value of the preset performance function. This indicates the maximum permissible steady-state error. Indicates the preset convergence rate, index number ; Define error transformation variables : (10)， Its derivative Represented as: (11)， In equation (11), Indicates error The first derivative, Indicates the preset performance function The first derivative.

9. The intelligent control method for preset performance of the electromechanical position servo system according to claim 8, characterized in that, Steps 2-3 are detailed below: According to the state-space equation, the position tracking error derivative Represented as: (12)， To avoid the differential explosion problem in the system, a virtual control input is designed for the controller. Design the following nonlinear dynamic surface control filter: (13)， In equation (13), Indicates the filtered signal The first derivative, This represents the positive damping parameter of the adjustable filter. Indicates the filtering error. Let represent a positive integrable function. express The upper bound estimate, of which express The first derivative, Updated by the following formula: (14)， In equation (14), express The derivative, Indicates a positive adaptive gain; Constructing barrier Lyapunov functions : (15)， right Differentiation yields : (16)， exist Then, substitute equations (11) and (12) into equation (16): (17)， Based on this, a virtual control law is designed. for: (18)， In equation (18), This represents the model's feedforward compensation term. Represents the linear robust feedback term. Indicates positive feedback gain; According to the state-space equation, the velocity tracking error derivative Represented as: (19)， exist Then, substitute equations (11) and (19) into equation (16): (20)， Based on this, the actual control law is designed. for: (21)， In equation (21), This represents the model's feedforward compensation term. Represents the linear robust feedback term. Represents the nonlinear robust term. This represents the feedforward compensation term in the RBF neural network. Indicates positive feedback gain; Design nonlinear robust terms for: (22)， In the formula, Describing the aggregate bounded disturbance Boundary value The estimated value, of which , , Represents a positive integrable function; Updated by the following formula: (23)， In the formula, express The first derivative, Indicates a positive adaptive gain; Proceed to step 3.

10. The intelligent control method for preset performance of the electromechanical position servo system according to claim 9, characterized in that, In step 3, the stability analysis of the designed intelligent controller with preset performance is performed using Lyapunov stability theory, and the asymptotic stability of the system tracking error is obtained, as follows: Constructing Lyapunov functions as follows: (24)， In equation (24), Both represent intermediate variables. The upper bound estimation error of the lumped disturbance is expressed as... , The first derivative of the virtual control law The upper bound of the estimation error, written as , The error in RBF weight estimation is written as: ; The stability of the designed intelligent controller with preset performance was analyzed using Lyapunov stability theory, and the asymptotic stability of the system tracking error was obtained.

Images