基于模糊逻辑的Markov链模型辨识方法

提出一种采用模糊Markov模型(FMM)，根据系统实际输入一输出建模的新方法，用于线性及非线性的随机动态系统辨识．讨论了动态系统的Markov链描述，给出了一阶受控Markov链的具体描述及其转移概率矩阵的计算方法；在传统模糊系统的基础上，探讨了建立FMM的方法，给出了FMM的仿真框图和建模的具体步骤；并给出了随机动态系统仿真的应用实例．基于模糊逻辑的Markov链模型解决了传统模糊系统不能处理随机现象的问题，同时也提高了Markov链模型辨识的速度．

An Identification Approach for Markov Chain Model Based on Fuzzy Logic

This paper focuses on the fuzzy-Markov modelling (FMM) approach for dynamic system. The problem under investigation is to improve a method for non-parametric modelling of linear as well as non-linear stochastic dynamic systems from input-output realiszations using fuzzy-Markov models. The paper introduced a kind of non-parametric Markov model and described its first order. The paper also discussed how to use FMM to identify stochastic system, and gave an example of stochastic dynamic systems' simulation for applications. The calculated simulation results for a dynamic system show preliminarily that the FMM can satisfy the need for identification of stochastic dynamic system.