非线性Hammerstein模型的辨识【毕业论文】.doc

本科毕业设计 （20 届） 非线性Hammerstein模型的辨识 所在学院 专业班级 电气工程及其自动化 学生姓名 学号 指导教师 职称 完成日期 年 月 摘　要 【摘要】Hammerstein模型是一种典型的非线性系统模型，其结构简单且能有效地描述常见的非线性系统的动态特性。本文采用改进的粒子群优化算法对该非线性系统进行辨识。改进的粒子群优化算法由粒子群优化算法（PSO）衍生、发展而来，PSO 算法是基于群体智能的一种的演化算法, 该算法来源于人工生命和演化计算理论，其显著的特点是算法中需要选择的参数少，程序实现简单，并在种群数量、寻优速度等方面和其他进化算法相比具有更大的优势。为了验证改进粒子群优化算法在系统辨识的精度和鲁棒性方面的优越性,文中通过Matlab数值仿真，讨论了改进型的PSO算法参数初值的设置与选择要求，并在不同信噪比下，和标准PSO算法和最小二乘法进行比较，从而在辨识精度,收敛速度等性能方面说明辨识的可靠性和有效性。 【关键词】系统辨识；Hammerstein模型；改进型PSO；最小二乘法 Abstract 【ABSTRACT】Hammerstein model is a typical nonlinear system,this model is simple and can effectively describe common characteristics of nonlinear dynamic system. In this paper, an improved particle swarm optimization applied for t nonlinear system identification.Improved PSO algorithm is developed from particle swarm optimization(PSO) algorithm,PSO algorithm is an evolutionary algorithm based on swarm intelligence and the idea comes from the theory of artificial life and evolutionary computation.The salient feature of PSO is not only the algorithm requires less parameters selected, the program simple but also have more advantages than other evolutionary algorithms in the population, searching speed and so on.In order to verify the excellence of the improved method in the precision and robust of the identification.The paper discusses the set and selection methods of initial parameters and selected requirement of the standard PSO algorithm through the Matlab simulation,.Then we compared with the traditional PSO algorithm and least square method, at different signal to noise ratio, to proof the effectivity and reliability of this algorithm in identification accuracy, convergence time. 【KEYWORDS】Hammerstein model；Identification；PSO；Method of least square 目　录 5 1 前言 6 1.1选题背景 6 1.2 Hammerstein模型研究现状 7 1.3 粒子群优化算法PSO算法 8 1.4 Matlab/ simulink简介 8 1.5 本章小结 9 2 Hammerstein模型 10 2.1Hammerstein模型的分