基于小波变换和神经网络的雷达滤波器故障诊断

Fault diagnosis of radar filter based on wavelet transform and neural network

由于模拟电路的非线性、容差性及元器件参数的连续可变性等特点的存在,传统的模拟电路故障诊断手段在实际运用中已经很难得到令人满意的结果。本文将小波变换与神经网络结合,对特征提取进行优化,以雷达滤波电路为研究对象,在考虑容差性的情况下对电路进行故障诊断,体现出小波神经网络在诊断正确性及时效性上所具有的优势。

With the consideration of actual analog circuit has some characteristics, such as nonlinear-ity, tolerance and the continuous variability of parameters on component, it is difficult to achieve expected results for traditional technology of analog circuit fault diagnosis in particular engineering. In this paper, wavelet transform and neural network are combined to optimization the feature extraction. Taking into account the tolerance, we researched fault diagnosis of radar filter. The result shows that wavelet neural network method reflects its adventages from both diagnosis correctness and timeliness.