A non-stationary signal correlator for on-line transit time estimation

The on-line reliable estimation and monitoring of transit time in BWRs is investigated. The interactive activation-competition artificial neural network is employed for providing an estimate of the current transit time directly at each incoming pair of signal values. As a result, transit time monitoring is accomplished for both constant and oscillating flow regimes, i.e. for both stationary and non-stationary signals. The proposed approach is robust to the presence of local and global component as well as to the addition of white uncorrelated noise.