基于知识库的电力变压器故障诊断专家系统  

Expert System of Power Transformer Fault Diagnosis Based on Knowledge Base

作　　者：刘守明[1] 胡志坤[2] 王美玲 

机构地区：[1]河南省电力公司周口供电公司,河南周口466001 [2]中南大学物理科学与技术学院,湖南长沙410083

出　　处：《计算机测量与控制》2011年第7期1569-1571,1574,共4页Computer Measurement &Control

摘　　要：针对电力变压器故障机理复杂多样的特征,提出了基于知识库为核心的电力变压器故障诊断专家系统,建立了以油中溶解气体分析(DGA)知识库、绝缘系统检测知识库及油中微水含量检测知识库三个知识库信息融合的故障确诊策略;其中,针对油中溶解气体分析知识库中知识量大的特征,使用规则置信度对一条故障信息可匹配多条规则知识的情况进行冲突消减;根据电力变压器故障种类概率呈现中间大两头小的特点,建立动态知识库模型以存入变压器近期故障,推理机进行推理时首先扫描动态知识库。在建立数据库的各个数据表时,通过设定主键和外键约束以使推理机推理更高效;实验结果表明,该专家系统在检测变压器故障时表现出了较高的准确率和较高的诊断速率。Aimed at the complex and the variety fault theory characteristic of the power transformer,expert system of power transformer based on knowledge base was proposed,and we put forward the fault diagnosis strategy based on three knowledge bases,including knowledge bases of Dissolved gas-in-oil analysis,measuring insulation system,and measuring water content.In view of the high knowledge content of the DGA knowledge base,we used rule confidence to avoid the collision for the fault signals which match multiply rules.Because the fault category of transformer has a character which probability is big at the centre but small in both end,we set up a dynamic knowledge base for the recent happened transformer fault,which was scanned firstly when the inference engine worked.Establishing primary key and foreign key when tables the datasheet of the database made the inference engine worked more efficiency;the result of the experiment indicates that the expert system has a high correctness and high efficiency.

关 键 词：电力变压器 知识库 推理机 

分 类 号：TM464[电气工程—电器]