模糊Petri网络知识表示方法及其在变压器故障诊断中的应用

Petri网络理论以图形化的方法直观地描述了离散事件系统的各种关系和行为，且以网络理论、代数理论等作为数学基础，适合于描述异步并发现象的计算机系统模型和对并行及并发系统进行行为分析的有效工具，在人工智能领域中正得到越来越广泛的应用。该文从模糊产生式系统出发，提出一种模糊Petri网络知识表示方法及其严密的推理算法，论述了利用模糊Petri网络(FPN)进行知识表示及推理的矩阵运算算法，为模糊知识的表示提供了一种有效的工具，并首次将其应用于电力变压器故障诊断中，“描述了故障征兆与故障的关系，使用了比传统专家系统更深的知识。与传统的专家系统相比，利用此理论及方法仅仅使用简单的矩阵计算，大大减少诊断时间，提高了准确度。最后通过实例对此算法进行了测试，表明该模型方法快速、准确。

APPLICATION OF FUZZY PETRI NETS KNOWLEDGE REPRESENTATION IN ELECTRIC POWER TRANSFORMER FAULT DIAGNOSIS

As a graph tool, Petri nets can be used to describe most of the relations and behaviors in the discrete event dynamic system. As a mathematical tool, the analysis of behavior properties and performance evaluation can be done conveniently, using net theory and algebraic theory. So it is applied in the field of artificial intelligence more and more widely. A method of Fuzzy Petri Nets(FPN) knowledge representation and its rigorous inference algorithm are proposed in this paper. Fuzzy knowledge representation and matrix operation algorithm of inference using Fuzzy Petri Nets are discussed. For the first time Fuzzy Petri Net is applied in electric power transformer, and it represents relations between fault symptoms and faults. Additionally, conventional expert systems applied to fault diagnosis need to manually build complicated knowledge databases, but Fuzzy Petri Net is very simple and clear. Because it only uses simple matrix calculation based on Petri nets theory, fast and accurate results are obtained. Finally the method of Fuzzy Petri Nets is tested by the application of some samples. It is proved that this method is very fast and accurate .