Abstract
Industrial process data have the characteristics of less label, multi-mode, high-dimension, and containing noises and outliers, which makes it costly to manually label data and increase the difficulty of mode identification and anomaly detection in process monitoring. To address the effect of these adverse factors, a semi-supervised discriminative projective dictionary pair learning (SSDP-DPL) is proposed for industrial process monitoring. Firstly, a semi-supervised dictionary pair learning framework with a class estimation regularization term is developed to address the problem of lacking labeled data. Based on unlabeled data and their reconstruction errors, the class estimation regularization term is designed to obtain a discriminative extended synthetical dictionary, mining the hidden discriminative information in unlabeled data and reducing the impact of incorrect class estimation. Secondly, a sparse constraint is added to the analytical dictionary to enhance the robustness of projective dictionary pair. Then, a class discriminative function term is designed to improve the discrimination of dictionary pair, which ensures the in-class compactness and inter-class separation of coding space. Finally, the synthetical dictionary, the analytical dictionary and the control threshold are obtained by iterating the dictionary pair to a process monitoring model for anomaly detection and mode identification. The effectiveness of the proposed process monitoring method is verified by continuous stirred tank heater process, Tennessee Eastman benchmark test. The proposed method is then applied to a real-world aluminum electrolysis industrial process. Experimental results demonstrate the superiority of our SSDP-DPL in contrast to other existing methods.
Funding Information
  • Key-Area Research and Development Program of Guangdong Province (2021B0101200005)
  • State Key Program of the National Natural Science of China (62133016)
  • National Natural Science Foundation of China (62003370)
  • Natural Science Foundation of Hainan Province (2021JJ30873)
  • Changsha Municipal Natural Science Foundation (kq2014137)

This publication has 35 references indexed in Scilit: