基于预警单元的逆变器并联系统分层故障诊断系统  

Parallel Connected Inverters System Fault Diagnosis Expert System Based on Early Warning Unit

作　　者：谢孟[1] 李耀华[2] 韩继文[1] 赵众[1] 张吉先 

机构地区：[1]中国科学院 [2]中国科学院电工研究所 [3]航天部3院33所

出　　处：《电气传动》2006年第8期38-42,共5页Electric Drive

摘　　要：随着逆变器并联系统规模的增大,电源系统对故障诊断提出了新的要求。结合逆变器并联系统故障诊断的特点,对所存在的问题进行了分析,并通过新的应用思路,将神经网络和专家系统应用到逆变器并联系统的故障诊断中。提出并定义了故障预警的一系列相关概念,设计了预警单元的工作结构。在此基础上提出了一种新的基于故障预警单元的分层故障诊断专家系统,详细阐述了故障预警单元的实现以及诊断系统的工作原理。通过原理分析及仿真结果可以看出,提出的方法对神经网络和专家系统在故障诊断中的有效应用具有很好的推动作用和一定的指导意义。With the increasing scale of the parallel connected inverters system, new requests of fault diagnosis are put forward by power supply systems. This paper analysis the character of the parallel connected inverters system fault diagnosis, then, apply the neural network and fuzzy control into the fault diagnosis. A series of concepts about fault early warning are defined. The framework of early warning unit is designed. Then a stratified fault diagnosis expert system based on early warning unit is proposed and the principle of fault early warning unit and diagnosis system are described in detail. Theory analyzing and some simulation results show that the method proposed in this paper will develop the practice use of neural network and expert system in fault diagnosis.

关 键 词：逆变器并联系统 故障诊断 专家系统 预警单元 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置] TP393.1[自动化与计算机技术—控制科学与工程]