ICS优化SVM在模拟电路故障诊断中的应用

Application of SVM Optimized ICS in Fault Diagnosis of Analog Circuit

针对容差模拟电路软故障,为了提高诊断的准确率,提出了一种基于改进布谷鸟算法优化支持向量机的故障诊断模型。首先,利用Hear小波分析对模拟电路进行故障特征提取;然后将提取的故障特征输入支持向量机进行故障诊断,同时为了使模型更稳定,利用改进布谷鸟算法选择最优惩罚参数和核函数参数以优化SVM。最后,以Sallen-key带通滤波器电路为例进行仿真实验,通过与神经网络、传统SVM分类模型进行对比,结果表明了该方法的优越性和可行性。

For the soft fault of analog circuit,a fault diagnosis model with improved cuckoo searchalgorithm(ICS)optimizing support vector machine(SVM)was proposed to improve the accuracy.First,Hear wavelet analysis was used to extract fault features as pretreatment from the signal of analog circuit.Then,the features were input to SVM to diagnose fault.And at the same time,ICS optimized parametersof SVM to improve the diagnosis accuracy.Finally,compared to neural network and SVM,Sallen-keyband-pass filter was taken as an example to prove the validity and feasibility of the proposed method.