Identification Method of Wiener Nonlinear System Based on Parameter Separation

Abstract

The invention discloses a method for identifying a Wiener nonlinear system based on parameter separation, comprising the following steps: A. converting a thermal power plant control system to be identified into a Wiener nonlinear system, and performing an input quantity of the thermal power plant control system to be identified Merge; B. Analyze the Wiener nonlinear system, including the linear dynamic part structure of the system, the nonlinear static part structure of the system, the dynamic disturbance type, and the measurement noise; determine n a , n b and n c , set the initial value, and collect the input data u(k) and y(k) repeatedly until k≥n a +n, k≥n b +n; C. Separate the time-invariant parameters and time-varying parameters of the Wiener nonlinear system; D. Identify the Wiener nonlinear system; E. When the identified model does not meet the requirements, return to step A and readjust the model structure and initial values, re-identify until a system model that meets the requirements is obtained. The invention can improve the deficiencies of the prior art, and improves the identification accuracy and convergence speed of the Wiener nonlinear system.

Description

technical field [0001] The invention relates to the technical field of industrial control, in particular to a parameter separation-based Wiener nonlinear system identification method. Background technique [0002] System identification technology is a branch of the control field. Its purpose is to use the input and output data of the system to identify the system parameter model and lay the foundation for system optimization and system control. System models are generally divided into linear models and nonlinear models, among which nonlinear models exist widely, and the identification of nonlinear models has attracted the attention of many engineers and scholars. The Wiener nonlinear system is the most typical nonlinear system, and its structure is composed of a linear dynamic part and a nonlinear static part connected in series. Therefore, Wiener nonlinear systems belong to the category of dynamic systems. The Wiener nonlinear system can describe most of the industrial sy...

AI Technical Summary

Problems solved by technology

However, this method is not suitable for situations where dynamic disturbances exist. Dynamic disturbances will be superimposed on the system output, reducing the identification accuracy of the system identification method, and even failing to identify the real parameters, which reduces the applicability of the method.

Scholars D.Wang, F.Ding, etc. in the literature "Least squares based and gradient based iterative identification for Wiener nonlinear systems", (simplified translation: the application of the least squares and gradient iterative algorithm in the identification of Wiener nonlinear systems, published in the International In the journal Signal Processing.91(5)(2011)1182-1189), two methods are proposed for Wiener nonlinear systems: the least squares iterative method and the gradient iterative method, which can use all input and output data at the same time. Under the condition of measurement noise, it has a good identification effect, but under the condition of dynamic interference, the identification accuracy will decrease

Furthermore, this method is an offline method, which utilizes all input and output data for each iteration, and has a large amount of calculation, which cannot be applied online.

Images