Hardware Implementation of Actuator Fault Estimation Applied to a Hydrographic Plant

Unknown input estimation have a major importance in real-time monitoring, diagnosis and fault-tolerance control. Consequently, this paper concentrates on state estimation for nonlinear system where a part of its inputs is unknown. In particular, the problem is oriented towards an experiential application for laboratory three tank system affected by actuator faults. Our objective is to perform the fault detection (FE) by the synthesis and the implementation of the unknown input observer (UIO). Simulation and experimental results confirm the performance of the suggested observer.