Fault Reconstruction for Continuous‐Time Systems Via Learning Observers

This paper addresses the problem of learning observer (LO)‐based fault reconstruction in continuous‐time systems. An LO is constructed such that simultaneous reconstruction of system states and actuator faults can be guaranteed. Then, existence conditions of LOs are first introduced to verify whether or not the LOs for fault reconstruction exist. An effective LO‐design algorithm is explored such that observer gain matrices can be conveniently solved using linear matrix inequality technique. An extension to sensor‐fault reconstruction is also investigated. Finally, two simulation examples are employed to demonstrate the effectiveness of the proposed LO‐based fault‐reconstructing approaches.