Low-complexity preset precision control method for full-state limited direct current motor system

Abstract

The invention belongs to the technical field of direct-current motor system control, and particularly relates to a low-complexity preset precision control method for a full-state limited direct-current motor system, which can embed state constraints and performance indexes into a controller design link by creatively fusing a dynamic constraint mapping function and a preset performance error conversion mechanism. The multi-target control requirement can be met only through a simple feedback structure, the calculation burden is remarkably reduced, and a new way is provided for efficient and reliable control of a direct-current motor system. The invention aims to provide a high-reliability and low-resource-consumption closed-loop control solution for a direct-current motor system, the full-state constraint and the preset dynamic performance are strictly met, the algorithm implementation complexity is remarkably reduced, and quick conversion of a theoretical result to industrial application is promoted.

Description

Technical Field

[0001] The invention belongs to the technical field of DC motor system control, and particularly relates to a low-complexity preset precision control method for a fully state-constrained DC motor system. Background Art

[0002] The significance of DC motor system control is to improve the efficiency, stability and reliability of the system, extend the service life of the motor, achieve high-precision control and reduce energy consumption. This has important application value in the fields of industry, automation, robotics and smart devices, and has promoted technological progress and industrial development. In the existing technology, for example: Reference 1: "Yang X, Deng W, Yao J. Neural network based output feedback control for DC motors with asymptotic stability [J]. Mechanical Systems and Signal Processing, 2022, 164: 108288."; Reference 1 designs a neural network-based output feedback controller to address the parameter uncertainty and external disturbance problems of the DC motor system. This method uses a neural network to approximate the unmodeled dynamics of the system online, and compensates for the disturbance through the output feedback mechanism, ultimately achieving stable tracking. The technical defects of Reference 1 include: (1) Failure to consider state constraints: No constraint protection is imposed on key state variables such as motor speed and current, and there is a risk of system damage due to physical limit violations. (2) High complexity: The neural network needs to update the weights online, which increases the computational burden, and the network structure design relies on empirical parameter adjustment, which limits its engineering practicality; Reference 2: "Xu Z, Xie N, Shen H, Hu X, Liu Q. Extended state observer-based adaptive prescribed performance control for a class of nonlinear systems with full-state constraints and uncertainties [J]. Nonlinear Dynamics, 2021, 105 (1): 345-358."; Reference 2 proposes a preset performance control method that combines an extended state observer (ESO) with an adaptive algorithm for full-state constrained nonlinear systems. The system's lumped uncertainty is estimated through ESO, and a preset performance function is introduced to constrain the tracking error boundary, ultimately achieving dynamic performance guarantee under full-state constraints. The technical defects of Reference 2 include: (1) Algorithm redundancy: The ESO observer and the adaptive law need to be run simultaneously, and the controller contains multi-layer coupled parameter update links, which has high computational complexity and is difficult to deploy on low-cost embedded platforms. (2) Conservative constraints: The preset performance function requires strict error bounds to be set in advance. If the initial state deviates significantly from the expected trajectory, it may cause the controller to saturate and fail.

[0003] To address the above deficiencies, there is an urgent need for a low-complexity control method that can achieve coordinated control of full-state constraints, preset performance, and input saturation compensation while avoiding online estimation algorithms. Therefore, the present invention proposes a low-complexity preset accuracy control method for a full-state constrained DC motor system. Summary of the Invention

[0004] The object of the present invention is to provide a low-complexity preset accuracy control method for a full-state constrained DC motor system. By innovatively integrating a dynamic constraint mapping function and a preset performance error conversion mechanism, state constraints and performance indicators can be embedded in the controller design process. Only a simple feedback structure is required to meet the multi-objective control requirements, significantly reducing the computational burden and providing a new approach for the efficient and reliable control of DC motor systems.

[0005] The technical solution adopted by the present invention is specifically as follows:

[0006] A low-complexity preset accuracy control method for a full-state constrained DC motor system includes the following steps:

[0007] Step 1: Describe the constrained problem:

[0008] For a DC motor system, it is represented by the following non-linear strict-feedback system

[0009]

[0010] where represents the system state vector, and represent the control gain and the additive system, and are the input and output of the system respectively, represents the unknown external disturbance of the system; the tracking error is defined as

[0011] e i = x i - α i-1 , i = 1, 2,..., n (2)

[0012] where is the tracking target, i = 1, 2,..., n - 1 are virtual control variables that need to be designed later; all states are subject to constraint limitations:

[0013]

[0014] where x ci > 0; and the control input is also subject to constraint limitations:

[0015]

[0016] where \(u\) M represents the maximum amplitude of the input; the goal of the controller is to achieve that the system tracking error satisfies the following preset performance under the condition that the system satisfies the state constraint and the input constraint:

[0017] \(\vert e\) i (t)\vert\lt\varepsilon\) i , \(t\geq T\) i (5).

[0018] Step 2: Then set the controller and perform low-complexity preset-precision control on the full-state constrained DC motor system through the controller.

[0019] In the said Step 2, the following steps are included:

[0020] Step 201: Input constraint processing; to effectively solve the problem of input constraint, the present invention uses a smooth function to approximate the input, and thus the input can be expressed as:

[0021] \(u(\tau)=\Pi(\tau,\theta)+\Delta u\) (6)

[0022] where

[0023]

[0024] and \(0\lt\theta\lt1\), \(\Delta u\) represents the difference between \(u\) and \(\Pi(\tau,\theta)\); according to the mean value theorem, \(\Pi(\tau,\theta)\) can be written as

[0025] \(\Pi(\tau,\theta)=\Pi(\tau_0,\theta)+\Pi\) v (\tau - \tau_0)\) (7)

[0026] where \(\tau\) v \(=k\tau+(1 - k)\tau_0\), and \(k\in[0,1)\); by choosing \(\tau_0 = 0\), we get

[0027] \(\Pi(\tau,\theta)=\Pi\) v \(\tau\) (8)

[0028] and \(0.42\leq\Pi\) v \(\leq1\); in summary, the control input can be expressed as

[0029] \(u(\tau)=\tau-\Lambda\tau+\Delta u\) (9)

[0030] and thus the original system formula (1) is expressed as

[0031]

[0032] Step 202: Definition of the preset performance function; design the preset performance function \(N\) i (t) in the following form:

[0033]

[0034] Step 203: Error transformation: To make the system meet the given preset performance constraints, define

[0035] δ i = e i (t) / N i (t) (12)

[0036] To ensure that δ i ∈ (-1, 1), introduce a new variable

[0037]

[0038] If it can be proven that ξ i is bounded, then δ i ∈ (-1, 1) can be satisfied.

[0039] Step 204: Virtual control and input controller design; To make the system meet the state constraints and input constraints, design the virtual control law:

[0040] α i = -k i sgn(g i )ξ i , i = 1, 2, …, n - 1 (14)

[0041] The input control law is designed as

[0042] τ = τ a + τ b (15)

[0043] where τ a = -k n sgn(g n )ξ n ,

[0044] The technical effects achieved by the present invention are:

[0045] In the present invention, the tracking accuracy and the finite-time convergence performance are improved

[0046] Experimental comparison: In the DC motor experimental environment, taking a DC motor with a rated voltage of 12V and a rated speed of 1500 rpm as the object, compared with the traditional PID, the method of the present invention shows advantages in the convergence time and accuracy: The present invention enables the speed tracking error to converge to the preset accuracy of ±16% within 3 seconds, while the accuracy of the traditional PID method has always remained at ±47%, thus showing the effectiveness of the present invention.

[0047] Theoretical support: The new preset performance function forces the error to enter the preset neighborhood within a finite time through a time-varying attenuation rate, avoiding the infinite-time convergence defect of the traditional PID control method.

[0048] In the present invention, input saturation compensation simplification and energy consumption reduction

[0049] Experimental comparison: Under the constraint of the input voltage limit of ±9V, compared with the traditional PID control method, the present invention shows advantages in input saturation compensation: during the entire control process, the control input of the present invention can maintain unsaturated control, while the control input of the traditional PID algorithm has been in a saturated state.

[0050] Theoretical support: The linear compensation term based on the mean value theorem of differentiation directly cancels the saturation nonlinearity, avoiding the additional energy loss of neural network approximation.

[0051] In the present invention, full-state constraint strict guarantee and robustness enhancement

[0052] Experimental comparison: When the motor rotor angle constraint is ±150 degrees and the rotational angular velocity constraint is ±1.5 degrees per second, compared with the traditional PID control method, the present invention shows advantages in the full-state overstep rate: under the method of the present invention, the rotor angle and angular velocity are always lower than the constraint limit values, while the traditional PID cannot guarantee it all the time.

[0053] Theoretical support: The dynamic constraint mapping embeds the state limit into the controller design through nonlinear transformation, avoiding the overstep risk from the mechanism. Brief description of the drawings

[0054] Figure 1 is a schematic diagram of the experimental platform structure of the DC motor system in the present invention;

[0055] Figure 2 is the control flow chart of the DC motor control system in the present invention;

[0056] Figure 3 is the trajectory of the system state x1 in the present invention;

[0057] Figure 4 is the trajectory of the system state x2 in the present invention;

[0058] Figure 5 is the trajectory of the system tracking error e1 in the present invention;

[0059] Figure 6 is the trajectory of the system tracking error e2 in the present invention;

[0060] Figure 7 is the trajectory diagram of the control input in the present invention;

[0061] Figure 1Among them, 1. Test bench base; 2. Motor driver; 3. DC motor; 4. Dynamic torque sensing; 5. Coupling; 6. Inertial load; 7. Rotary encoder; 8. Air switch. Specific implementation mode

[0062] In order to make the purpose and advantages of the present invention clearer, the present invention will be specifically described below in conjunction with embodiments. It should be understood that the following text only describes one or several specific implementation manners of the present invention, and does not strictly limit the scope of protection specifically requested by the present invention.

[0063] As Figure 1 shown, for the low-complexity preset-precision control method of the full-state constrained DC motor system, aiming at multiple complex problems such as parameter uncertainty, external disturbance, unmodeled dynamics, and input saturation in the full-state constrained DC motor system, as well as the deficiencies of existing control methods in terms of computational complexity, co-optimization of state constraints and preset performance, and simplification of anti-saturation compensation, the present invention proposes a low-complexity preset-precision control method, aiming to achieve the following goals:

[0064] (1) Break through the co-control bottleneck of state constraints and preset performance: Without relying on neural networks, adaptive algorithms or complex observers, through a new control architecture, strictly ensure the physical constraints of all state variables such as DC motor speed and current at the same time, and achieve that the output tracking error converges to the preset precision within a finite time (such as dynamic performance indicators such as specified convergence time and overshoot), overcoming the limitation that traditional methods can only achieve asymptotic stability or infinite-time convergence.

[0065] (2) Significantly reduce the computational complexity of the controller: Abandon the redundant structures based on online parameter estimation, multi-module coupling or auxiliary system design in the existing technology, and construct a lightweight feedback controller through the direct embedding of a dynamic constraint mapping and error conversion mechanism, meeting the stringent requirements of industrial scenarios for real-time performance and embedded deployment.

[0066] (3) Simplify the input saturation compensation design: Adopt a fusion strategy of smooth function approximation and differential mean value theorem, and directly offset the nonlinear influence of input saturation with a single compensation term, avoiding the tuning difficulty and computational burden introduced by complex neural networks or piecewise control structures in existing anti-saturation methods, and improving engineering applicability.

[0067] Through the above technical breakthroughs, the present invention is committed to providing a high-reliability and low-resource-consumption closed-loop control solution for DC motor systems. While strictly meeting the full-state constraints and preset dynamic performance, it significantly reduces the algorithm implementation complexity and promotes the rapid transformation of theoretical results into industrial applications.

[0068] The present invention specifically includes the following steps:

[0069] Step 1: Describe the restricted problem:

[0070] For a DC motor system, it is represented by the following non - linear strict - feedback system

[0071]

[0072] where represents the system state vector, and represent the control gain and the additive system, and are the input and output of the system respectively, represents the unknown external disturbance of the system; define the tracking error as

[0073] e i = x i -α i-1 , i = 1, 2, …, n (2)

[0074] where is the tracking target, i = 1, 2, …, n - 1 are virtual control variables that need to be designed later; all states are subject to constraint limitations:

[0075]

[0076] where x ci > 0; and the control input is also subject to constraint limitations:

[0077]

[0078] where u M represents the maximum amplitude of the input; the goal of the controller is: under the condition that the system satisfies the state constraints and input constraints, make the system tracking error satisfy the following preset performance:

[0079] |e i (t)| < ε i , t ≥ T i (5).

[0080] Step 2: Then set the controller and perform low - complexity preset - accuracy control on the fully - state - restricted DC motor system through the controller.

[0081] In Step 2, the following steps are included:

[0082] Step 201: Input - restriction processing; to effectively address the problem of input restriction, the present invention uses a smooth function to approximate the input, and thus the input can be expressed as:

[0083] u(τ) = Π(τ, θ)+Δu (6)

[0084] wherein

[0085]

[0086] and 0 < θ < 1, Δu represents the difference between u and Π(τ,θ); according to the mean value theorem, Π(τ,θ) can be written as

[0087] Π(τ,θ) = Π(τ0,θ) + Π v (τ - τ0)(7)

[0088] wherein τ v = kτ + (1 - k)τ0, and k ∈ [0,1); by choosing τ0 = 0, we get

[0089] Π(τ,θ) = Π v τ(8)

[0090] and 0.42 ≤ Π v ≤ 1; in summary, the control input can be expressed as

[0091] u(τ) = τ - Λτ + Δu(9)

[0092] Furthermore, the original system formula (1) is expressed as

[0093]

[0094] Step 202: Definition of the preset performance function; design the preset performance function N i (t) in the following form:

[0095]

[0096] Step 203: Error transformation: To make the system satisfy the given preset performance constraint, define

[0097] δ i = e i (t) / N i (t)(12)

[0098] To ensure that δ i ∈ (-1,1), introduce a new variable

[0099]

[0100] If it can be proved that ξ i is bounded, then δ i ∈ (-1,1) can be satisfied.

[0101] Step 204: Design of virtual control and input controller; In order to make the system satisfy the state constraints and input constraints, design the virtual control law:

[0102] α i =-k i sgn(g i )ξ i , i = 1, 2, …, n - 1 (14)

[0103] The input control law is designed as

[0104] τ = τ a +τ b (15)

[0105] where τ a =-k n sgn(g n )ξ n ,

[0106] In the present invention, based on the smooth approximation of input saturation and the differential mean value compensation technique

[0107] Core of the technique: A composite saturation input approximation method based on the smooth hyperbolic tangent function and the differential mean value theorem is proposed.

[0108] Implementation method: Use a continuously differentiable smooth function to directly approximate the input saturation non - linear characteristic; Combine the differential mean value theorem, convert the difference between the saturated input and the ideal input into an equivalent disturbance term, and design a single linear compensation term to be embedded in the controller, avoiding an auxiliary system or a multi - layer parameter update structure.

[0109] Advantages: Compared with the traditional complex anti - saturation strategies in the literature (such as gated recurrent neural network, piece - wise control), the form of the controller is significantly simplified, the real - time calculation amount is reduced, and no additional parameter tuning is required.

[0110] In the present invention, the co - design of the finite - time preset performance function and state constraints

[0111] Core of the technique: Construct a new type of time - varying preset performance function and dynamic constraint mapping mechanism to achieve the cooperative control of full - state constraint and finite - time convergence of the tracking error.

[0112] Implementation method: First, introduce a time - varying function with finite - time decay characteristics to strictly constrain the upper and lower bounds of the tracking error, ensuring that the error converges to the preset accuracy (such as the zero - error neighborhood) within the user - specified time. Second, map the original constrained state to an unconstrained virtual state through a non - linear transformation term, and design the control law based on the backstepping method to make the actual state always satisfy the preset physical constraints (such as speed and current limit).

[0113] Advantages: Break through the limitation that the traditional BLF method only guarantees asymptotic stability or infinite-time convergence.

[0114] In the present invention, a low-complexity backstepping controller architecture

[0115] Technical core: Propose a backstepping control framework without an online estimation algorithm, and reduce the computational complexity through structural innovation.

[0116] Implementation method: First, discard the neural network weight update, adaptive law or extended state observer, and directly use the system state feedback to construct the control quantity. Second, uniformly embed the input saturation equivalent disturbance compensation, state constraint mapping, and preset performance error conversion into the recursive design process of the backstepping controller to form a single closed-loop control law.

[0117] Advantages: Compared with the control method of a multi-layer adaptive structure, the computational amount is greatly reduced, which is suitable for low-computing-power embedded platforms.

[0118] In the present invention, (1) methodological innovation: For the first time, fuse the smooth saturation approximation-differential mean value compensation and the finite-time preset performance-state constraint mapping to form a unified control paradigm of "low-complexity architecture + strong constraint guarantee"; (2) technical breakthrough: Without the premise of an online estimation algorithm, realize the triple-objective coordination of full-state constraint, finite-time convergence, and input saturation compensation, filling the gap in the traditional method that it is difficult to balance "complexity-performance".

[0119] In the present invention, the tracking accuracy and the finite-time convergence performance are improved

[0120] Experimental comparison: In the DC motor experimental environment, taking a DC motor with a rated voltage of 12V and a rated speed of 1500 rpm as the object, comparing with the traditional PID, the method of the present invention shows advantages in the convergence time and accuracy: The present invention makes the speed tracking error converge to the preset accuracy of ±16% within 3 seconds, while the accuracy of the traditional PID method has always remained at ±47%, thus showing the effectiveness of the present invention.

[0121] Theoretical support: The new preset performance function forces the error to enter the preset neighborhood within a finite time through a time-varying decay rate, avoiding the infinite-time convergence defect of the traditional PID control method.

[0122] In the present invention, the input saturation compensation is simplified and the energy consumption is reduced

[0123] Experimental comparison: Under the constraint of the input voltage limit of ±9V, comparing with the traditional PID control method, the present invention shows advantages in the input saturation compensation: The control input of the present invention can maintain unsaturated control throughout the control process, while the control input of the traditional PID algorithm has always been saturated.

[0124] Theoretical support: The linear compensation term based on the differential mean value theorem directly cancels the saturation nonlinearity, avoiding the additional energy loss in the approximation of the neural network.

[0125] In the present invention, the full-state constraint strictly guarantees and enhances the robustness.

[0126] Experimental comparison: When the motor rotor angle constraint is ±150 degrees and the rotational angular velocity constraint is ±1.5 degrees per second, compared with the traditional PID control method, the present invention shows advantages in the full-state overstep rate: under the method of the present invention, the rotor angle and angular velocity are always lower than the constraint limit values, while the traditional PID cannot guarantee it all the time.

[0127] Theoretical support: The dynamic constraint mapping embeds the state limit into the controller design through nonlinear transformation, avoiding the overstep risk from the mechanism.

[0128] To verify the effectiveness of the present invention, a DC motor system experimental platform as Figure 1 shown was built. This platform includes the following components: Actuator: DC motor 3, motor driver 2, dynamic torque sensor 4, rotary encoder 8, inertial load 6, coupling 5; The DC motor 3, motor driver 2, dynamic torque sensor 4, rotary encoder 7, inertial load 6, coupling 5 are correspondingly installed on the experimental bench base 1, and two air switches 8 are installed on the experimental bench base 1. Specifically, the DC motor 2 uses Panasonic MHMF042L1U2M, the motor driver 3 uses Panasonic MBDLT25SF, the model of the dynamic torque sensor 4 is: DYN-200, and the model of the rotary encoder 7 is Omron E6B2-CWZ1X; Support unit: Experimental bench base 1, power supply system with air switch 8; Control system: Real-time control software and monitoring software based on the Advantech PCI-1723 development board, and the system framework is as Figure 2 shown. The experimental sampling interval is set to 0.05 seconds, and the control algorithm is deployed to the PCI-1723 hardware platform through the real-time control software to achieve the closed-loop control of the motor speed and torque. The monitoring software is responsible for data acquisition, status visualization and online analysis of performance indicators.

[0129] To highlight the effectiveness of the present invention, we will compare it with the traditional PID control method. In the following experiments, we consider the cases where the tracking target is of low frequency and high frequency at the same time, and compare the mean (A), standard deviation (S) and maximum value (M) of the tracking error and torque.

[0130] Case 1: Given a target trajectory y of the top frequency r =100 + 30sin(t / 3.14)°, the traditional PID controller is designed as where the parameter is designed as K p =0.6, K i =0.05, Kd = 3. The control parameters of the present invention are designed as k1 = 15, k2 = 10, T1 = 5s, T2 = 5s, A0 = 110, A1 = 0.1, ε1 = 20, ε2 = 0.5. The corresponding results are as Figures 3 - 7 shown, and the control performance is shown in Table 1.

[0131] Controller PID method Proposed method <![CDATA[A e1 > 17.7135 5.8737 <![CDATA[S e1 > 26.7387 7.5903 <![CDATA[M e1 > 54.8822 14.3679 <![CDATA[A e2 > 1.4200 0.4455 <![CDATA[S e2 > 2.0134 0.7556 <![CDATA[M e2 > 1.9506 0.9558 <![CDATA[A u > 8.2339 0.0832 <![CDATA[S u > 11.7404 0.1055 Mu 9.0000 0.1634

[0132] Table 1: Comparison Table of Control Performance

[0133] From Figures 3 - 7 and Table 1, it can be seen that: The comparison results between the target trajectory yr and the system state x1 are as Figure 3 shown, and the trajectory of the state variable x2 is as Figure 4 shown. It can be seen that although there is a certain gap between the target trajectory and the system state due to the limitations of the physical system, the proposed controller has better performance compared with the PID method. In addition, when the proposed method is adopted, the amplitudes of all states are kept within the constraint range, while the PID method cannot achieve this.

[0134] The changing trajectories of the tracking errors e1 and e2 are respectively as Figure 5 and Figure 6 shown. It can be observed that the proposed controller can control both e1 and e2 within the preset boundaries, while the PID method cannot, which verifies the effectiveness of the preset performance strategy. The trajectory of the control input signal is as Figure 7 shown, and the control input amount when the proposed method is adopted is significantly smaller than that of the PID method, which indicates that the proposed input constraint handling scheme has a significant effect.

[0135] The above is only the preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and retouches can be made, and these improvements and retouches should also be regarded as the protection scope of the present invention. The structures, devices, and operation methods not specifically described and explained in the present invention, unless otherwise specified and limited, are implemented according to the conventional means in the art.

Claims

1. A low-complexity preset-precision control method for a fully state-constrained DC motor system, characterized in that: Including the following steps: Step 1: Describe the restricted problem: For a DC motor system, it is represented by the following non-linear strict feedback system wherein represents the system state vector, and represent the control gain and the additive system, and are the input and output of the system respectively, represents the unknown external disturbance of the system; Define the tracking error as e i = x i - α i-1 , i = 1, 2, …, n (2) Among them is the tracking target, i = 1, 2, …, n - 1 are virtual control variables; all states are subject to constraint limitations: where x ci > 0; and the control input is also subject to constraints: where \(u\) M represents the maximum amplitude of the input; the objective of the controller is: under the condition that the system satisfies the state constraints and input constraints, to achieve that the system tracking error meets the following preset performance: |e i (t)| < ε i , t ≥ T i (5). Step 2: Then set the controller, and perform low-complexity preset accuracy control on the fully state-restricted DC motor system through the controller.

2. The low-complexity preset-precision control method for the fully state-constrained DC motor system according to claim 1, wherein: In the said Step 2, it includes the following steps: Step 201: Input restriction processing; approximate the input using a smooth function, and then the input can be expressed as: u(τ) = Π(τ,θ) + Δu (6) Where And 0 < θ < 1, Δu represents the difference between u and Π(τ,θ); according to the mean value theorem, Π(τ,θ) can be written as Π(τ,θ) = Π(τ0,θ) + Π v (τ - τ0)(7) where τ v = kτ+(1 - k)τ0, and k ∈ [0,1); by choosing τ0 = 0, we get Π(τ,θ) = Π v τ (8) 0.42 ≤ Π v ≤ 1; In summary, the control input can be expressed as u(τ) = τ - Λτ + Δu (9) Then the original system formula (1) is expressed as 3. The low-complexity preset-precision control method for a fully-state-constrained DC motor system according to claim 1, characterized in that: The said Step 2 also includes the following steps: Step 202: Definition of the preset performance function; design of the preset performance function N i (t) is in the following form:

4. The low-complexity preset-precision control method for the fully-state-constrained DC motor system according to claim 1, wherein: The said Step 3 also includes the following steps: Step 203: Error transformation: To make the system meet the given preset performance constraints, define δ i = e i (t) / N i (t) (12) To ensure that δ i ∈ (-1, 1), a new variable is introduced If it can be proven that ξ i is bounded, then δ i ∈ (-1, 1) can be satisfied.

5. The low-complexity preset-precision control method for the fully state-constrained DC motor system according to claim 4, characterized in that: The said Step 2 also includes the following steps: Step 204: Virtual control and input controller design; to make the system meet the state constraints and input constraints, design the virtual control rate: α i = -k i sgn(g i )ξ i , i = 1, 2, …, n - 1 (14) The input control rate is designed as τ = τ a + τ b (15) where τ a = -k n sgn(g n )ξ n ,

Images