摘要
提出了基于小波多分辨分析和小波包预处理的模拟电路故障诊断方法。该方法用小波作为信号预处理工具,经小波多分辨分析得到N层分解后的低频和高频信号,再利用小波包分析对多分辨分析没有细分的高频信号进一步分解,以达到提高频率分解率的目的。经PCA分析和归一化后的能量作为训练样本送入BP神经网络进行训练。仿真实验表明此方法能够快速有效的对模拟电路的故障进行诊断和定位。
A method for fault diagnosis of analog circuits based on wavelet multi-resolution analysis and wavelet packet transform is presented. Using the wavelet decomposition as a preprocessor, extracted the feature information by wavelet de-noising. The collected data was processed by wavelet multi-resolution analysis to draw the features in low frequency and high frequency single, then used wavelet packet transform to improve frequency resolution. The normalization energy is finally used to train a BP neural network to diagnose faulty components in an analog circuit. Simulation results illustrate the method for fault diagnosis is quickly and effectively.
出处
《电脑与信息技术》
2014年第6期22-25,共4页
Computer and Information Technology
基金
晓庄学院第三批产学研合作示范基地项目(项目编号:20140616-1)
关键词
多分辨分析
小波包变换
BP神经网络
故障诊断
multi-resohition analysis
wavelet packet transform
BP neural network
fault diagnosis
