矿用干式变压器局部放电模式识别方法

Partial discharge pattern recognition method for mine-used dry-type transformer

为提高矿用干式变压器局部放电模式识别准确率,提出了一种矿用干式变压器局部放电模式识别方法。首先,采用正交匹配追踪算法对原始局部放电信号进行去噪,最大程度保留原始局部放电信号的有用信息;然后,通过自回归模型提取去噪后局部放电信号的自回归系数特征;最后,将自回归系数特征输入随机森林集成分类器对局部放电模式进行识别。实验结果表明,该方法平均识别准确率达98%。

In order to improve recognition accuracy of partial discharge pattern of mine-used dry-type transformer,a partial discharge pattern recognition method for mine-used dry-type transformer was proposed.Firstly,orthogonal matching pursuit algorithm is used to denoise original partial discharge signal,which can retain useful information of the original partial discharge signal to the greatest extent.Then,autoregressive coefficient features of the partial discharge signal after denoising are extracted by autoregressive model.Finally,the autoregressive coefficient features are input into random forest integrated classifier to recognize partial discharge pattern.The experimental result shows that average recognition accuracy of the method is98%.