遗传算法在变压器故障诊断中的应用

根据变压器油中溶解气体分析的三比值法建立了基于Mamdani模糊推理的变压器故障诊断系统。该故障诊断系统由历史数据库、模糊推理诊断模块和优化模块三部分组成。利用遗传算法和历史试验数据对模糊推理模块的隶属度函数进行优化．克服了传统的三比值法对变压器进行故障诊断时存在的临界值判据缺损问题，并可对故障诊断过程中的编码缺损情况作出解释。实例验证，该诊断系统具有较好的诊断能力，可提高变压器故障诊断的准确性。

Fuzzy diagnosis system optimized with genetic algorithm for power transformer

Based on the three-ratio method of dissolved gas analysis for power transformer, a fuzzydiagnosis system obeying Mamdani fuzzy reasoning rules is presented. The system composes of history database,fuzzy reasoning and diagnosing module,optimization module. Its fuzzy membership functions are optimized by genetic algorithm and historical test data,the lack of criterion for critical value in three-ratio method is overcome and the lacking codes in fault diagnosis can be explained. Ten sampling data are diagnosed by the proposed system,and the results show that it improves the veracity of transformer fault diagnosis. This project is supported by the National Key Technologies R & D Program of China(2002BA218C).