一、前言 本文(上)曾谈到,建立一个过程的数学模型是实现过程自动控制的关键一环。不仅工程技术方面如此,其它如地质、气象、经济计划、企业管理、医学生物等方面的定量的研究工作也要涉及模型识别的问题。随着四个现代化的需要,这一问题的研究将越来越深入、广泛。所谓模型识别,按Zadeh的定义是“在已给输入与输出的基础上,从某类系统中选定一个系统,它与所测系统是等价的”。即从某类系统M={S}中选定它的一个元素So,它刻画某一输入与输出的关系作为模型。将同样的输入u加到被测系统与模型L,它们的输出y与总相 I. Preface This article (above) mentions that establishing a mathematical model of a process is a key part of achieving process automation. Not only in terms of engineering, but other quantitative research work such as geology, meteorology, economic planning, business management and medical biology also involves the problem of model identification. With the needs of the four modernizations, the research on this issue will be more and more in-depth and extensive. The so-called model identification, according to Zadeh’s definition is “on the basis of the input and output, select a system from a certain type of system, it is equivalent to the measured system.” That is, a certain element So is selected from a certain type of system M = {S}, which describes a certain input-output relationship as a model. The same input u is added to the system under test and the model L, and their output y and the total phase