模糊系统论

Theory of Fuzzy Systems

本文从讨论系统观出发,给出了一般模糊系统的形式定义和表示方法,全面讨论了模糊系统的静态结构、动态行为、学习机制、数字特征及问题求解过程等问题。论文还讨论了这种模糊系统的表示能力问题,证明了诸如逻辑电路系统、Petri网、H网、各种数据驱动的系统、用流程图表示的程序系统等都是这种系统的特例。论文提出的模糊系统及其“浸润式”执行机制为新型计算机系统的设计提供了一种新模式。一个物理的模糊系统可用一个由大量微处理器构成的互联网络来实现。为此,论文提出了一些具体实现这种系统的设想以及在已有计算机上进行模拟的具体设计。

Started from an overview of systems, a formal definition of fuzzy systems is presented, and then a series of relevant issues about fuzzy systems, such as statical structure, dynamic behavior, learning mechanism, numerical characteristics of the fuzzy systems and problem solving by the fuzzy systems is discussed comprehensively. The paper also discusses the expressive power of fuzzy systems and proves that many existing systems, such as logical circuit systems, Petri net, H net, various data driven systems, program systems described by flowcharts, all can be represented by fuzzy systems as some special cases. The system execution mode by 'infiltration' presented by the paper can be used to guide the design of a new kind of computer systems which is composed of a massive micro-processors connected by a fuzzy net. Finally, a simulation scheme of the new computer system on the existing computers and imagination about implementation of the system are also presented.