In most chemical processes, both online measurements and offline laboratory analysis can be obtained at different sampling rates. Usually, the online process variables are frequently sampled while the key quality indi...
详细信息
In most chemical processes, both online measurements and offline laboratory analysis can be obtained at different sampling rates. Usually, the online process variables are frequently sampled while the key quality indicators are analyzed at irregular time, sometimes on hourly and sometimes daily basis. To effectively integrate different classes of measurements in a multi-rate process, a multi-rate probability principal component analysis (MPPCA) model is proposed to utilize the efficiently collected data and to improve the performance on both model prediction and process monitoring. In MPPCA, the model parameters are calibrated by the expectation-maximization algorithm. The proposed method is able to handle irregular samples in measurements and incorporate all the observations for model training. Also, the corresponding statistics based on MPPCA is developed for the fault detection purpose. Finally, a TE benchmark is presented to illustrate the effectiveness of the proposed methods. (C) 2017, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.
暂无评论