基于模糊数学和Markov链的三比值法及其应用

Fuzzy Mathematics and Markov Chain Based Three Ratio Method and Its Application

为了解决三比值法区间边界过于绝对化和编码缺失的问题,提出了基于模糊数学和Markov链的三比值法。基于模糊数学和Markov链的三比值法利用模糊数学的方法模糊边界值,并利用Markov链进行时间序列的进一步处理,使诊断结果更加准确。通过采用改进前、后的两种三比值法对40组油色谱数据进行检验,结果表明基于模糊数学和Markov链的三比值法解决了传统三比值法边界过于绝对化的缺点,同时很好地处理了编码缺失问题,对于变压器故障诊断的结果准确率更高,也更符合故障发展的规律。

In order to solve problems of too absolute section boundary of three ratio method and code missing,a fuzzy mathematics and Markov chain based three ratio method is proposed.In fuzzy mathematics and Markov chain based three ratio method,the boundary value is fuzzied up by using the fuzzy mathematics method,and further treatment of time sequence is done by using Markov chain which makes diagnosis results more accurate.Two kinds of three ratio methods before and after the modification are used to testify 40 sets of oil chromatographic data.Results show that the fuzzy mathematics and Markov chain based three ratio method solves the too absolute section boundary defect of traditional three ratio method and deals with the code missing problem very well,which has higher result accuracy rate of transformer fault diagnosis and more complies with the development rules of fault.