Process decomposition and distributed fault detection of large-scale industrial processes

In this work, a data-driven distributed process monitoring problem is considered. A process decomposition approach is proposed by leveraging community structure detection to decompose variables into subsystems optimally via finding a maximal value of the metric of modularity. A two-layer distributed monitoring scheme is developed where local estimators are designed based on the configured subsystems of variables using canonical correlation analysis. Inner-subsystem interactions and inter-subsystem interactions are tackled by the two layers separately. The proposed method is applied to a simulated wastewater treatment process.