基于多基因GEP的非线性系统辨识方法  

A Method of Nonlinear System Identification Based on Multi-gene GEP

作　　者：白焰[1] 朱耀春[1] 蒋毅恒[1] 

机构地区：[1]华北电力大学自动化系,北京102206

出　　处：《微计算机信息》2008年第13期267-269,共3页Control & Automation

摘　　要：提出了一种利用基因表达式编程(GEP)的多基因特性进行非线性系统辨识的方法,可根据NARMAX模型结构特点形成模型结构,比原有GP算法更方便和高效。提出了可变终端集策略,通过仿真试验说明了可变终端集在模型搜索搜索方面比固定终端集具有更快的收敛速度和更大的适应性。A new system identification method based on gene expression programming （GEP）, which makes use of the muhigene characteristic of GEP and can obtain the mode structure according to the NARMAX modes＇ expression, is provided in this article, and has a higher convenience and efficiency than GP, A strategy of variable terminal sets is introduced, and the simulation results express that the variable terminal sets has a higher convergence speed and a higher adaptability than the fixed terminal sets.

关 键 词：非线性系统 系统辨识 NARMAX模型 基因表达式编程 

分 类 号：TP15[自动化与计算机技术—控制理论与控制工程]