主元分析法在MIMO系统中的故障检测与诊断研究

摘要

Using principal component analysis (PCA) realized fault detection and diagnosis system for multiple-input multiple-output (MIMO) object .It was possible to overcome the shortcomings of the mechanism analysis diagnosis .The association between the variables was eliminated ,and the effects of ambient noise was reduced .The experiments on the multi-variable level control show the PCA method can monitor the production process effectively ,and diagnose the faults in the system accurately and timely .%针对多输入多输出对象应用多元分析方法实现系统的故障检测与诊断，能够有效克服机理分析法诊断的缺点，既可以消除各变量之间的关联，也减小外界噪声的影响。通过对多变量液位控制进行试验验证，主元分析法能够有效地对生产过程进行监测，并准确及时地分析判断系统中发生的故障。