中文摘要：

MSPCA方法在生产过程监控方面有着广泛应用。本文在研究该方法的基础之上，提出了一些改进，在其进行小波分解后即对其小波系数进行阈值处理，使小波消噪与MSPCA方法合为一体，并运用统计控制图中的平方预测误差（SPE）图方法检测引起过程变化或故障的过程变量。在保证其MSPCA算法复杂度不变的前提下，能够消除数据的噪声污染，使故障诊断的误报大为减少。经检验，该算法确实可行，相对于小波消噪与MSPCA方法分别进行，效率提高了大约13％-17％。

英文摘要：

MSPCA method has wide application on the process monitoring. On the basis of studying MSPCA, this paper introduces an improvement method that will threshold the wavelet coefficients when the data is decomposed, and it can combine the wavelet rectification with the MSPCA, and the squared prediction error （SPE） of the statistical control graph is used to detect the process variants which induce the change or fault of the process. On the principle of invariability of the complexity, it can eliminate the contamination of the data, and decrease the false alarm of the fault diagnosis. As the test, this method is feasible. Comparing with wavelet rectification and MSPCA which carry on respectively, the efficiency raised about 13%- 17%.