支持向量数据描述(support vector data description,SVDD)是一种单值分类方法,该方法能够在缺少故障样本的情况下,通过采集到的正常状态数据样本建立起单值分类器,从而区分出机器的运行状态。笔者提出了一种小波变换支持向量数据描述...支持向量数据描述(support vector data description,SVDD)是一种单值分类方法,该方法能够在缺少故障样本的情况下,通过采集到的正常状态数据样本建立起单值分类器,从而区分出机器的运行状态。笔者提出了一种小波变换支持向量数据描述诊断方法,利用正交小波变换(orthogo-nal wavelet transformation,OWT)方法提取各细节信号的峰峰值作为输入参数,用SVDD方法进行分类。通过滚动轴承的试验数据分类结果显示,该方法比没有提取特征值的SVDD可有效处理复杂机械振动信号,提高了诊断的准确率。展开更多
A new method which is caIled orthogonal wavelet transform was introduced into deconvolution voltammetry to deal with discrete data containing strong random noise. From simulated data and experimental data, the process...A new method which is caIled orthogonal wavelet transform was introduced into deconvolution voltammetry to deal with discrete data containing strong random noise. From simulated data and experimental data, the processed results show that wavelet trans form would be the best method in processing complex discrete signal containing high white noise,and was compared with Fourier least square method. Wavelet transform has many merits than Fourier method.展开更多
文摘A new method which is caIled orthogonal wavelet transform was introduced into deconvolution voltammetry to deal with discrete data containing strong random noise. From simulated data and experimental data, the processed results show that wavelet trans form would be the best method in processing complex discrete signal containing high white noise,and was compared with Fourier least square method. Wavelet transform has many merits than Fourier method.
