论文部分内容阅读
异常识别是多元统计过程控制(MSPC,Multivariate Statistical Process Con-trol)方法有效应用的关键.针对现有研究对历史异常信息利用的不足,综合考虑了主成分变量贡献率与重构误差变量贡献率对异常识别的影响,将两种变量贡献率进行归一化处理并求和得到综合变量贡献率;提出了一种基于综合变量贡献率的MSPC异常识别方法,并基于matlab计算平台实现了该算法.通过田纳西过程故障模式仿真及异常识别,对该方法的应用及算法有效性进行了实例验证.
Anomaly identification is the key to the effective application of MSPC (Multivariate Statistical Process Con-trol) method.According to the shortcomings of the existing research on the utilization of historical anomalous information, the contribution rate of principal component variables and the contribution rate of reconstruction error variables For the effect of anomaly identification, the contribution rates of two variables are normalized and summed up to obtain the contribution rate of synthetic variables. A MSPC anomaly identification method based on the contribution rate of synthetic variables is proposed, and the algorithm is implemented based on the matlab calculation platform Through the Tennessee process failure mode simulation and abnormal identification, the application of the method and the validity of the algorithm are validated by examples.