Petri网络在变压器故障诊断中的应用

Petri网络理论以网络理论，代数理论等为数学基础，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，适用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，在人工智能领域中正得到越来越广泛的应用。作者提出一种基于Petri网络的变压器故障诊断模型，论述了利用Petri网络进行知识表示及推理的矩阵运算算法。描述了故障征兆与故障的关系。使用了比传统系统更深的知识、简单，明确。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，可以减少诊断时间，提高准确度。通过实例对此模型进行的测试表明该模型方法是快速、准确的。

APPLICATION OF PETRI NETS IN TRANSFORMER FAULT DIAGNOSIS

The Petri nets theory which takes the net theory and algebra theory for mathematical basis can be used to describe the various relations and behaviors in the discrete event dynamic system by graphical means. Because the Petri nets is suited to describe the computer system model under the asynchronous concurrent phenomena and to make the behavior analysis on parallel and concurrent system, so it is more and more widely applied in artificial intelligence field. In this paper a fault diagnosis model for power transformer based on Petri nets is proposed in which by use of Petri nets the matrix operation algorithm for knowledge representation and inference is dissertated, the relation between the fault omen and the fault itself is described. The presented model is simple and clear, in which only simple matrix calculation based on Petri nets theory is used. Thus, the time for fault diagnosis can be obviously reduced and the diagnosis accuracy is improved. Testing the presented model by practical examples, the results show that this model is fast and accurate.