Search

This paper focuses on the fault-tolerant control problem for an Internet-based three-tank system in the presence of possible sensor bias faults. The Internet-based three-tank system is an experimental setup that can be regarded as a typical networked system for evaluating networked fault-diagnosis and fault-tolerant control methods. Packet dropout phenomenon in the sensor-to-controller link is considered in this paper, and the fault type we deal with is chosen as the sensor bias fault. Fault-diagnosis unit is designed toward an auxiliary system. Sensor bias faults can be detected by comparing the residual signal generated by the fault detection filter and a prescribed threshold. After that, the fault can be isolated by using the residual analysis approach. Once the fault is isolated, it can be estimated iteratively in the least-squares sense. A fault accommodation method is proposed, and a fault-tolerant control strategy is achieved based on the fault information provided by the fault-diagnosis unit. The approach brought forward in this paper is demonstrated via an experimental study on the practical Internet-based three-tank system. Results show the effectiveness and the applicability of the proposed techniques. [ABSTRACT FROM PUBLISHER]

This brief is concerned with the active fault-tolerant control (FTC) problem for an Internet-based networked three-tank system (INTTS) serving as a benchmark system for evaluating networked FTC algorithms. The INTTS has two parts located at Tsinghua University in China and at the University of South Wales in the U.K., respectively, which are connected via the Internet. With the INTTS as an experimental platform, the active FTC problem is investigated for a class of nonlinear networked systems subject to partial actuator failures. Once a specific actuator failure is detected and confirmed by a fault diagnosis unit, the control law is then reconfigured based on the information of the detected fault. Both the stability and the acceptable $H_\infty $ disturbance attenuation level are guaranteed for the closed-loop system using the remaining reliable actuators. Extensive experiments are carried out on the active FTC problem of the INTTS with partial actuator failures, and the effectiveness of the proposed scheme is illustrated. [ABSTRACT FROM AUTHOR]