论文部分内容阅读
提出了一种基于小波包去噪和主元分析的故障检测和诊断方法。该方法利用小波包分解系数收缩的信号去噪法先对正常工况下的数据进行处理,然后运用T2统计、Q统计方法,结合主元得分图和变量贡献图对一模型进行了仿真研究。结果表明,该方法是有效的。
A fault detection and diagnosis method based on wavelet packet denoising and principal component analysis is proposed. The method uses the signal denoising method of wavelet packet decomposition coefficient contraction to process the data under normal working conditions first, and then uses T2 statistic and Q statistic method to combine the principal component score graph and variable contribution graph to simulate a model. The results show that the method is effective.