In this research, predictor-based system identification is recursively implemented to identify system in closed-loop, which has been connected with proper threshold setting for fault detection. The predictor-based system identification technique is based on the assumption that the process and measurement noise is white. This paper shows how recursive system identification is connected with fault detection, which can remove the influence of noise. The effectiveness is demonstrated for recursive system identification combined with proper threshold setting via Matlab simulation.

This content is only available via PDF.
You do not currently have access to this content.