Canonical Probabilistic Principal Component Analysis based Operating Quality Monitoring for Industrial Control Systems

In the present day, industrial control systems (ICS) provide a significant contribution to the improvement of production efficiency. In the case of industrial control systems, the node arrangements are massive, security vulnerabilities are inevitable, and the operational status is complex, posing a threat to the stability of industrial processes. This study proposes a canonical variate analysis with probability principal component analysis monitoring method for ICS. Operational data are considered the most important information. The first step is to select significant variables that affect operating performance using oneway variance analysis. Through the application of probabilistic principal component analysis on the constantly linked canonical subspace and its residual subspace, a model for simultaneous monitoring is established. Thereafter, monitoring statistics and control limits based on online data are established. An analysis of the actual drilling process confirms the efficacy and superiority of the proposed strategy. The technology has superior sensitivity relative to traditional monitoring techniques.