基于最小二乘法的系统参数辨识

The System Parameter Identification Based on Least Squares

系统辨识是自动控制学科的一个重要分支,由于其特殊作用,已经广泛应用于各种领域,尤其是复杂系统或参数不容易确定的系统的建模。过去,系统辨识主要用于线性系统的建模,经过多年的研究,已经形成成熟的理论。但随着社会、科学的发展,非线性系统越来越受到人们的关注,其控制与模型之间的矛盾越来越明显,因而非线性系统的辨识问题也越来越受到重视,其辨识理论不断发展和完善。本文重点介绍了系统参数辨识中最小二乘法的基本原理,并通过热敏电阻阻值温度关系模型的辨识实例,具体说明了基于最小二乘法参数辨识在Matlab中的实现方法。结果表明基于最小二乘法具有算法简单、精度较高等优点。

System identification is an important branch of automatic control,because of its special role,has been widely used in various fields,especially in complex systems or parameter is not easy to determine the system modeling. In the past,system identification was mainly used to model linear systems,and after years of research,mature theories have been formed. But with the development of society,science,nonlinear system more and more get the attention of people,the contradiction between the control and model more and more obvious,thus the problem of nonlinear system identification is paying more and more attention,the continuous development and improvement of the recognition theory. Article mainly introduced the basic principle of system parameter identification method of least squares,and through the thermistor resistance temperature relation model identification instance,specify the parameter identification based on least squares method in Matlab. The results show that the least squares method possesses the advantages of simple algorithm and higher precision.