Control System Identification Methods Research And Application

In order to solve frequent manual control,low system stability and low automatic control rate,scholars have proposed many advanced control technologies,but most of the modern technologies are based on exact objects mathematical model.System identification is a kind of effective method to establish a mathematical model of the process object through measuring input and output data of the system.In this thesis,the theory and application of the least square method and the particle swarm optimization algorithm for the identification of eight kinds of common discrete system models are studied.Through lots of simulations with MATLAB,the applicability of the two algorithms is verified and it is concluded that the least squares identification accuracy is superior to the conclusion of particle swarm optimization algorithm under conditions of step response.At the same time,this paper give a selection of inertia weight and swarm size witch make particle swarm optimization a better performance for discrete system identification.Then,using the actual collection of input and output data from industry,this thesis makes some research on discrete system identification with the two algorithms.Through simulation results analysis,the applicability of the two algorithms is verified and it is concluded that particle swarm optimization algorithm identification accuracy is superior to the conclusion of the least squares under conditions of no step response.