基于范例推理的变压器油中气体分析综合诊断模型  

Synthetic Diagnosis Model Applying Case-Based Reasoning in Dissolved Gas Analysis of Transformer Oil

作　　者：钱政[1] 高文胜[1] 尚勇[1] 严璋[1] 

机构地区：[1]西安交通大学电气工程学院,710049

出　　处：《电工技术学报》2000年第5期42-47,57,共7页Transactions of China Electrotechnical Society

基　　金：国家自然科学基金;东北电力集团资助项目

摘　　要：以变压器油中溶解气体的分析结果为基础 ,建立基于范例推理的变压器绝缘故障综合诊断模型。先用三种不同的范例检索算法分别检索出与待诊变压器故障征兆相近的源范例集 ,然后经对检索的源范例进行综合判断获得和待诊变压器故障征兆最相近的最佳源范例 ,并用其信息得到待诊变压器的诊断结论。实例检验结果表明 ,该模型比现有方法有更高的诊断可靠性和实用前景。A synthetic diagnosis model applying case-based reasoning in dissolved gas analysis of transformer oil is utilized to diagnose the fault of power transformer.First,three kinds of retrieving models are adopted to retrieve the set of source case which the symptom is similar to the symptom of suspected transformer,then an algorithm of synthetic estimation is introduced to achieve the optimal one of source cases,the diagnosis result of suspected transformer could be acquired by the comparison of symptoms between the optimal case and the suspected transformer at last.The application results show that the purposed model improves the reliability and practicality.

关 键 词：变压器油 气体分析 范例推理 电力变压器 

分 类 号：TM855[电气工程—高电压与绝缘技术] TM410.7