Stirred-Tank Heating System Fault Diagnostic Mechanism Based Mahalanobis Distance

Predictive maintenance of the plants can be performed using multivariate sensor data gathered from the manufacturing and process sectors. This data represents actual operation behaviors. The intricate behaviors of industrial systems, sensor interactions, control system corrections, and variability in aberrant behavior make anomaly identication and diagnosis—a crucial component of predictive maintenance— to be more and more challenging. Specic chemical processes necessitate extra stringent requirements in addition to high-precision actuator functioning. Even slight changes in the outcome product's quality can result from chemical interactions. Thus, in addition to the requirement for a high-performance integrated control system, monitoring operations must be quick and accurate enough to identify and isolate defects when system issues arise. This research investigates a data-driven estimation based process fault diagnostic and detection approach. According to this approach, the discrepancy between the process response and the process model response is used to identify the process failure. For fault diagnostic purposes, errors are classied using the Mahalanobis distance. The technique is validated in this study using the stirred-tank heating process. The outcomes of the simulation show how effective the suggested strategy is.