摘要
本文分析了系统辨识中常用的几种离散系统模型的结构,这些离散系统模型及其预测器模型在形式上都能表示成线性回归模型,并且一个复杂的模型能由几个简单的模型所构成。根据离散系统模型结构的这些特点,我们用面向对象的程序设计思想,用“类”的概念来实现各种离散系统模型,从而形成离散系统模型类库,并可由此派生出其它更具针对性的模型类。此离散系统模型类库和递推辨识算法类库结合在一起,可方便地构成完整的系统辨识程序。本文将介绍这一离散系统模型类库的层次结构及模型类库在系统辨识中的一些应用,给出两个例子。
The implementation of discrete system models has been studied in this paper. Discrete system models in common use can be expressed in linear regression form and a complex system model can be expressed in combination of several simple system models. In accordance with object-oriented programming concepts,discrete system models are implemented with 'class'and a class library has been explored.lt is found that, with this discrete system model class library, programs for application in system identification can be built conveniently.Two examples are given.
出处
《计算机应用与软件》
CSCD
北大核心
2001年第1期43-49,共7页
Computer Applications and Software
基金
国家自然科学基金(编号:69774002)
关键词
系统辨识
离散系统
继承
派生
模型类库
面向对象
System identification Discrete system Model Class Inheritance Derivation
