Fault detection of nuclear reactors by estimation of unknown input for systems with disturbances

Some traditional methods for parameter estimation and fault detection may lead to poor results with nuclear systems in practice because approximations and uncertainties in the system models possibly result in unexpected plant unavailability. This experience has led to an interest in development of estimation and fault detection methods for nuclear systems. In this paper, we present a framework for estimation of friction force during refueling. Specifically, an unknown input Kalman filter is introduced for reducing the uncertainty associated with the data, that arise due to many reasons including random noise and disturbances. Based on our proposed model, the estimation technique developed by this paper is used to estimate the friction force during condition monitoring of the graphite core of Advanced Gas Reactor.