基于随机森林算法的模拟电路故障诊断

Fault diagnosis of analog circuits based on Random Forest algorithm

摘要:

为提高模拟电路参变故障的诊断率，提出基于多特征向量提取和随机森林(RF)算法的模拟电路故障诊断新方法。采用时域和频域特征向量组合的多维特征向量以反映不同故障特征，经RF算法进行决策，并对决策树棵数及候选特征向量个数进行优化。故障诊断实验结果表明，所提方法能较好地实现容差模拟电路故障诊断，与支持向量机(SVM)方法相比，表现出更好的分类性能；与小波(包)特征提取方法相比，简化了多维数据特征提取步骤，易于实现在线故障诊断。

Abstract:

In order to improve the diagnostic rate of analog circuit parametric fault, a new fault diagnosis method based on multi-feature extraction and Random Forest(RF) algorithm is presented. By combining the time domain and frequency domain feature vectors for multi-feature vector data to reflect different faults, the decision is made by RF algorithm, the number of decision tree and candidate feature vectors are optimized. The fault diagnosis experiment results show that the proposed method can better realize analog circuit fault diagnosis, and shows a better classification performance compared with Support Vector Machine(SVM) method, and compared with wavelet(packet) feature extraction method, the multi- dimensional data feature extraction is simplified for online fault diagnosis.