Skip to Main Content
This paper presents a fault detection and diagnosis method for unknown systems with unknown input. Using the same formulation as described in , a residual signal is formed which is independent of the unknown input but dependent on the unknown system parameters and the measurements. An identification algorithm with an improved modification procedure for the covariance matrix is obtained and is then used to track parameter changes caused by system faults. The method is demonstrated through its application to a hydraulic turbine generating unit.