CN-116224788-B - Model reference self-adaptive disturbance rejection optimization control method

CN116224788BCN 116224788 BCN116224788 BCN 116224788BCN-116224788-B

Abstract

The invention discloses a model reference self-adaptive disturbance rejection optimization control method, which comprises the steps of generating a first control signal by a model reference self-adaptive method by using a model reference system, obtaining a second control signal generated by a PD controller after disturbance compensation in an actual system, adding the first control signal and the second control signal to be used as total control signals, controlling a controlled object by using the total control signals to obtain an actual system output, inputting the actual system output and the total control signals into an extended state observer in the actual system to obtain an estimated amount of the total disturbance, and then carrying out feedback compensation on the actual system total disturbance by using the estimated amount of the total disturbance, and constructing an Lyapunov function based on the error by using an error between a reference model output and the actual system output in the model reference system to obtain a self-adaptive law, so as to update the first control signal, achieve the effects of system self-adaption and output error reduction, and enable the system to be more stable and have stronger disturbance rejection capability.

Inventors

WEI WEI
LI WEI
WU HONGHAO

Assignees

北京邮电大学

Dates

Publication Date: 20260512
Application Date: 20230117

Claims (4)

1. The model reference adaptive disturbance rejection optimization control method is characterized by comprising the following steps of: generating a first control signal by a model reference adaptive method using a model reference system ; In the actual system, a second control signal generated by the PD controller through disturbance compensation is acquired ; First control signal And a second control signal Adding as a total control signal By means of a total control signal Controlling the controlled object to obtain the output of the actual system ; Output the actual system And total control signal Inputting into an extended state observer in an actual system to obtain an estimated quantity of total disturbance Reuse of Performing feedback compensation on the total disturbance of the actual system; using reference model outputs in model reference systems And actual system output The error between the first control signals is calculated, a Lyapunov function based on the error is constructed, and an adaptive law is obtained, so that the first control signals are updated The method specifically comprises the following steps: updating the first control signal using the second formula The second formula is: , In the formula, As a function of the gain parameter(s), , All are positive definite matrixes, and the sum of the positive definite matrixes, Is a vector of a constant row and, Output for reference model And actual system output And a column vector formed by the errors and the first order derivative of the errors.
2. The model reference adaptive disturbance rejection optimization control method according to claim 1, wherein the actual system output The first formula is satisfied, and the first formula is: , In the formula, Is that Is used for the first derivative of (c), As a result of the overall perturbation, As an estimate of the total disturbance, Is a scaling factor for the PD controller, Is the differential coefficient of the PD controller, Is set to be the set value, For the estimated value of the actual system output, Outputting first order leads for actual system Is used for the estimation of the (c), As a function of the gain parameter(s), As a result of the overall control signal, Is the first control signal.
3. The model reference adaptive immunity optimization control method according to claim 1, wherein, Determining according to a third formula, wherein the third formula is as follows: , In the formula, For a state vector of a model reference system, Is a state vector of the actual system.
4. The model reference adaptive disturbance rejection optimization control method according to claim 1, wherein the first control signal is generated by a model reference adaptive method using a model reference system The method specifically comprises the following steps: Giving a first control signal according to a fourth formula The fourth formula is: , In the formula, In order to adjust the parameters of the device, Is a gain parameter.

Description

Model reference self-adaptive disturbance rejection optimization control method Technical Field The invention discloses a model reference self-adaptive disturbance rejection optimization control method, and belongs to the technical field of disturbance rejection control. Background Active disturbance rejection control (Active Disturbance Rejection Control, ADRC) is a control technique based on the standard type of feedback system (integrator series type) proposed at the beginning of the disfigurement of classical control theory, targeting the robustness of engineering control. The concept is that the PID control which is dominant in the industry is taken as a starting point, the active disturbance rejection concept is provided on the basis of improving the nonlinear PID, the algorithm is simple, and the control precision can be still maintained under the actions of unknown strong nonlinearity and uncertain strong disturbance. Active disturbance rejection control currently mainly comprises three aspects of a nonlinear tracking differentiator, an extended state observer (Extended State Observer, ESO) and a series of active disturbance rejection control law designs. ESO can estimate not only the state, but also the "total disturbance", making ADRC a great advantage in dealing with non-linearities, uncertainties and disturbances. Adjusting the controller bandwidth and observer bandwidth can achieve a satisfactory control result. However, for ADRC, the actual system bandwidth is often limited, and high frequency noise is also introduced by the high bandwidth, and when the bandwidth is limited, the ESO estimation is inaccurate, so that a larger estimation error is brought, and the system dynamics is affected. Disclosure of Invention The application aims to provide a model reference self-adaptive disturbance rejection optimization control method to solve the problems that the dynamic state of a system is not ideal when the bandwidth in the existing disturbance rejection control technology is limited, high-frequency noise is introduced due to high bandwidth, and the like. The invention provides a model reference self-adaptive disturbance rejection optimization control method, which comprises the following steps: Generating a first control signal u ad by a model reference adaptive method by using a model reference system; Acquiring a second control signal u' adrc generated by the PD controller through disturbance compensation in an actual system; adding the first control signal u ad and the second control signal u' adrc to obtain a total control signal u, and controlling a controlled object by using the total control signal u to obtain an actual system output y p; inputting the actual system output y p and the total control signal u into an extended state observer in the actual system to obtain an estimated amount of the total disturbance Reuse ofPerforming feedback compensation on the total disturbance of the actual system; And constructing an error-based Lyapunov function by utilizing the error between the reference model output y m and the actual system output y p in the model reference system to obtain an adaptive law, so as to update the first control signal u ad. Preferably, the actual system output y p satisfies a first formula, which is: In the formula, Is the second derivative of y p, f is the total disturbance,For the estimated amount of the total disturbance, k p is the proportionality coefficient of the PD controller, k d is the differential coefficient of the PD controller, r is the set value, z 1 is the estimated value of the actual system output, and z 2 is the first order derivative of the actual system outputB 0 is a gain parameter, u is a total control signal, and u ad is a first control signal. Preferably, updating the first control signal u ad specifically includes: updating the first control signal u ad using a second formula: Where B 0 is the gain parameter, k c＝Γ1+Γ2+Γ3,P、Γ1、Γ2、Γ3 is the positive definite matrix, B kT∈R2 is the constant row vector, e R 2 is the reference model output y m and the actual system output y p, and the column vector is composed of the first order of the errors. Preferably, e is determined according to a third formula: e=xm-xp where x m is the state vector of the model reference system and x p is the state vector of the actual system. Preferably, the first control signal u ad is generated by a model reference adaptive method by using a model reference system, which specifically includes: The first control signal u ad is given according to a fourth formula: Where k 1、k2、k3 is an adjustable parameter and b 0 is a gain parameter. Compared with the prior art, the model reference adaptive disturbance rejection optimization control method has the following beneficial effects: The invention achieves the satisfying effect which can not be achieved by a single ADRC through a Control strategy of combining Model REFERENCE ADAPTIVE Control (MRAC) with active disturbance rejection