Skip to Main Content
This paper presents the application results concerning the fault detection of a dynamic process using linear system identification and model-based residual generation techniques. The first step of the considered approach consists of identifying different families of linear models for the monitored system in order to describe the dynamic behaviour of the considered process. The second step of the scheme requires the design of output estimators (e.g., dynamic observers or Kalman filters) which are used as residual generators. The proposed fault detection and system identification schemes have been tested on a chemical process in the presence of sensor, actuator, component faults and disturbance. The results and concluding remarks have been finally reported.