基于连续小波奇异熵的模拟电路故障诊断新方法

Novel approach for analog circuit fault diagnosis based on continuous wavelet singularity entropy

针对模拟电路的故障诊断和定位问题,为进一步提高故障诊断准确率,提出了一种基于连续小波Tsallis奇异熵和超限学习机的故障诊断方法。首先应用连续小波变换计算被测电路时域响应信号的时频系数矩阵,然后将其分割为8个相同大小的子矩阵,分别计算每个子矩阵的Tsallis奇异熵,组成特征向量,最后将特征应用于超限学习机多类分类器进行区分。仿真结果表明,故障诊断方法能较好地获取故障响应信号的本质特征,并具有较其他现存方法更高的故障诊断正确率。

Aiming at the issue of analog circuit fault diagnosis and location, a novel approach for analog circuit fault diagnosis based on continuous wavelet Tsallis singularity entropy （TSE） and extreme learning machine （ELM） is proposed to enhance the accuracy of fault diagnosis. Firstly, the fault response signals are preprocessed by the continuous wavelet transformation to obtain the time-frequency coefficient matrix, and the matrix is divided into 8 congruent time-frequency blocks. Then, the feature vector is obtained by computing TSE of each block. Finally, the feature vectors are used as the inputs of a kind of muhiclass classifier, namely ELM. The simulation results demonstrate that the proposed fault diagnosis approach can not only extract the essential features of fault response signals with better performance, and also achieve higher diagnosis accuracy than other reported approaches.