Robust fault detection for chemical processes

This paper deals with the design of a robust fault detection filter for a two stage chemical reactor which is a complex process, susceptible to faults, that can minimize productivity and humans safety. This process is considered as a time-delay system with both unknown inputs and parameter uncertainties. Therefore, a multi-objective optimization technique is used to introduce a new performance index and formulate the robust fault detection filter design problem as a H_∞ model-matching problem. Stability conditions are verified using linear matrix inequality (LMI). Adaptive thresholds are later developed in order to evaluate fault indicators. Simulation tests show the efficiency of the proposed approach.