To obtain the informative enough data from the modern chemical processes, various sensors and measuring instruments have been applied. Among them, both the online measurements and offline laboratory analyser results are collected and used for multivariate statistical process monitoring purpose. Hence, it indicates that the obtained measurements indeed contain different sampling rates due to the different demands of the control systems and the equipment constraints. To handle the multi-rate process monitoring problem, it is desirable to effectively integrate the measurements with different sampling rates. In this paper, a multiple probability principal component analysis (MPPCA) model is proposed for efficient collection of data and the improvement of the performance on model prediction and process monitoring. In the proposed model, it has been derived as a general form for any multi-rate systems and the model parameters are calibrated by the expectation–maximization algorithm. Based on it, several statistics are constructed for different sampling rates in the process monitoring scenario. Finally, two case studies are presented to show the effectiveness of the proposed method.

为了从现代化学过程中获得足够的信息，已应用了各种传感器和测量仪器。其中，在线测量结果和离线实验室分析仪结果均被收集并用于多变量统计过程监视目的。因此，它表明，由于控制系统的不同要求和设备限制，所获得的测量结果确实包含不同的采样率。为了处理多速率过程监视问题，期望有效地将测量与不同采样速率集成在一起。本文提出了一种多概率主成分分析（MPPCA）模型，以有效地收集数据并提高模型预测和过程监控的性能。在建议的模型中，它已被推导出为任何多速率系统的通用形式，并且模型参数已通过期望最大化算法进行了校准。基于此，在过程监视场景中针对不同的采样率构造了一些统计信息。最后，提出了两个案例研究，以证明该方法的有效性。