Kalman filter based FDI of networked control system

Abstract

Networked control system (NCS) deals with how to realize the close-loop control through network. Fault diagnosis for NCS is a new research region. In this paper, network channel connects the sensor and the controller. So NCS was regarded as a system with output time-variant delay and white noise. The iterative method was used to compensate the time-delay. Then different Kalman filters were constructed for NCS according to different system faults. So the respective residues were generated and a fault detection and isolation (FDI) approach was introduced. At last, an example was given to show the feasibility of the approach.