Fault Estimation and Fault Tolerant Control for Network Connected Systems via Observer Design

This paper focuses on fault estimation and fault tolerent control for network connected systems with sensor fault. First, disturbance observer are used for distubrance isolation for single linear system with sensor faults. Then, leaderless network connected systems are considered and consensus protocols are designed for the systems based on the disturbance and state estimations. The stability conditions are derived with the aid of Lyapunov functions. In addition, based on our design, no matter how big the sensor fault is, the consensus can still be achieved and the consensus errors are ultimatedly bounded. Finally, simulations are provided and the simulation results prove the effectiveness of the design in this paper.