Fault diagnosis for networked control systems with delayed measurements and inputs

This study researches the fault diagnosis approach for networked control systems (NCSs) with delayed measurements and inputs. Different from current fault diagnosis approaches that regard all delays in NCSs as an input delay, this paper presents a new fault diagnosis approach for NCSs that considers delayed measurements and inputs simultaneously and presents a discrete delay-free transformation approach for delayed measurements. By constructing an augmented system that comprises the states of faults and by delay-free transformations of delayed measurements and inputs for the augmented system, the problem of fault diagnosis for NCSs with delayed measurements and inputs is transformed into the problem of state observation of a delay-free augmented system. A new design approach to the fault diagnoser is presented for NCSs by constructing a novel reduced-order state observer of the augmented system, which can real-time diagnose faults without using residual to embody faults. Finally, the feasibility and effectiveness of the proposed approaches are validated by a numerical example.