摘要
对控制系统进行监视很有必要。本文提出一种控制系统性能的监视方法。首先通过采集控制系统正常运行时的响应曲线数据得到参考数据块，然后对此严重相关的数据块进行主元分析。主元分析是解决数据相关并降低数据维数的一种非常有效的多元统计方法，在主元分析中经常用到的统计量是Ｔ２ 统计量，Ｔ２ 统计量是用来检验新观测值是否仍落在主元确定的控制域内。根据主元分析得到的结果，利用Ｔ２ 统计量，提出一种新的性能指标，用来判断控制系统运行的状态是否正常，从而使操作人员能够方便地对控制系统进行监视。仿真结果表明，该方法是一种有效的监视控制系统性能的方法。
The control system monitoring is important. An approach to monitor the performance of the control system is proposed. Principal component analysis(PCA) is used for the preliminary analysis of data sampled from the control system in normal operation. PCA is a very efficient multivariate statistical method to treat the autocorrelated data and reduce the data dimensions. Based on the result of PCA, an index of the control system performance is introduced to determine whether the control system is normal or not. Thus it is convenient for operators to monitor the control system. Simulation result shows that the method is effective.
出处
《数据采集与处理》
CSCD
1999年第3期330-333,共4页
Journal of Data Acquisition and Processing
