基于小波变换和CFA-LSSVM模拟电路故障诊断

为了提高模拟电路软故障诊断、识别的正确分类率,提出了一种提升小波变换和混沌萤火虫算法(CFA)优化LSSVM参数的模拟电路故障诊断方法。首先对采集到的被测电路输出电压信号进行提升小波变换;然后对变换后的数据进行因子分析法对优化处理,将经优化的数据作为不同模式的故障特征集;最后将所得故障特征集作为样本输入到CFA-LSSVM模型进行故障诊断。实验结果表明,该方法的故障诊断正确率达到了98%以上,提高了诊断性能,可适用于模拟电路的故障诊断。

Analog circuit fault diagnosis based on wavelet transform and CFA-LSSVM

In order to improve the correct classification rate of analog circuit fault diagnosis and recognition,a simulation circuit fault diagnosis method based on lifting wavelet transform and chaotic firefly algorithm (CFA) is proposed to optimize LSSVM parameters.Firstly,the wavelet transform is applied to the output voltage signal of the measured circuit.Then,the transformed data is analyzed by factor analysis method,and the optimized data is taken as the fault feature set of different modes.Finally,the obtained fault feature set as sample is imported into the CFA-LSSVM model for troubleshooting.The experimental results show that the fault diagnosis accuracy of this method is more than 98％,which improves the diagnostic performance and can be applied to the fault diagnosis of analog circuits.