论文部分内容阅读
大系统是一种变量甚多的多变量系统。把通常的多输人多输出系统参数估计的方法用于大系统,则往往因维数太高,使计算量过大,而难于实现。大系统参数估计必须克服这种困难。主要有两个途径:一是简化,二是分散。本文综述了采用这两个途径的现有方法。主要是摄动方法、借助于分布式计算机系统可用于在线估计的递阶结构方法以及用较小计算机就可离线进行大系统参数估计的递阶结构方法。综述着重于阐明主要概念,指出现有方法的优缺点和发展方向。
A big system is a multivariable system with many variables. When using the method of estimating the parameters of a normal multi-input multi-output system in a large system, it is often difficult to implement the method because the number of dimensions is too high and the amount of calculation is too large. Large system parameter estimates must overcome this difficulty. There are two main ways: First, simplify, and second, decentralize. This article reviews existing approaches that use both approaches. Mainly perturbation method, with the help of distributed computer system can be used for on-line estimation of the hierarchical structure method and the smaller computer can be off-line large system parameter estimation of the hierarchical structure method. The review focuses on clarifying the main concepts and pointing out the advantages and disadvantages of the existing methods and the direction of their development.