基于不同RS与NN组合的数据挖掘配电网故障诊断模型

在将基于RS（粗糙集）理论的数据挖掘用于配电网故障定位诊断模型研究的基础上，提出和构造了4类不同的RS与NN（神经网络）组合的故障诊断模型，给出了RS与NN在4类模型中实现不同的互补性，关联关系，应用机理和原则及相应的局限性。通过对5类模型的仿真测试结果比较，证明数据挖掘模型潜在知识发现的重要意义，并对4类模型中RS的应用机理和性能作出全面的评估。文中对RS和数据挖掘研究的评估对其他领域的故障诊断研究具有同样的指导意义。

ASSESSMENT OF DATA MINING MODEL BASED ON THE DIFFERENT COMBINATION ROUGH SET WITH NEURAL NETWORK FOR FAULT SECTION DIAGNOSIS OF DISTRIBUTION NETWORKS

The study is based on rough set theory(RST) and data mining (DM) in use for fault section diagnosis(FSD)of distribution networks. The four kinds of diagnosis models on the basis of the different combination RS with neural network (NN) are proposed. The different complementarity and associated relation, the different application mechanism and principle, the model limitation are presented by implementing RS and NN in four models and the all around assessment of the application mechanism and performance of RS is given. Furthermore, comparisons of simulation results among the five kinds of models prove the significance of DM's discovery in hidden knowledge. The conclusion and assessment of each model have the same significance for fault diagnosis system in other fields. This project is supported by National Natural Science Foundation of China (No. 59877016).