Parameter Identification Of Wiener Model Based On Intelligent Optimization Algoithm

Nonlinear systems exit widely in practice. It is a basis to build a mathematical model for analyzing and controlling nonlinear system precisely. Modeling a nonlinear system includes determining the structure and estimating the parameters of it. Since it can't find a general structure to describe various nonlinear systems, block-oriented models are widely adopted by researchers to describe the system. Wiener model is a kind of a block-oriented model, which consists of a linear dynamic subsystem followed by a static nonlinearity. In practice, the nonlinearity is versatile and the intermediated signal is unavailable, that are difficult problems to identify Wiener model. In this thesis, intelligent optimization algorithms are used to identify Wiener model. The main works of this thesis are as follows:1. The parameter identification of Wiener model with discontinuous nonlinearity is considered. The sum of squares of output error between the actual system and the estimated system is used as the objective function and the hybrid particle swarm optimization is adopted to search the parameters such that the objective function is minimized. The proposed method can estimate the parameters of Wiener model simultaneously and don't need the intermediated signal.2. For a class of Wiener model with complex nonlinearity, we assume that the nonlinearity is reversible and a two-segment polynomial is used to approximate the inverse of the nonlinearity. Then, the hybrid particle swarm optimization is used to estimate the parameters of the system.3. A novel method is proposed to deal with the unavailability of the intermediated signal. Step signal is feed to the Wiener model, which separates the nonlinearity from Wiener model. The linear dynamic subsystem and the nonlinearity of Wiener model are identified in two steps using particle swarm optimization. 4. A fuzzy Wiener model (FWM) is proposed. In the proposed FWM, ARMR model is used to describe the linear dynamic subsystem and the nonlinearity is represented by a Takaki-Sugeno (T-S) fuzzy model. A self-adaptive differential evolution algorithm is used to estimate the parameters of the linear part and the T-S fuzzy model simultaneously.5. Finally, the identification and control of the pH neutralization process are studied. The T-S fuzzy model is used to represent the titration curve and its inverse. The two-step method is used to estimate the parameters. Based on the identified model, a PID controller is designed, where the parameters of PID controller is tuned by differential evolution algorithm.