Simultaneous Estimation of Neutron Flux and Reactivity in Nuclear Reactors

In this article, we propose a methodology based on cubature Kalman filtering (CKF) for simultaneous estimation of two important variables of nuclear reactors, viz. the neutron flux and the total core reactivity, from signals of delayed response self-powered neutron detectors (SPNDs). Moreover, by using the linearity of our model, the Rao-Blackwellized CKF (RBCKF) algorithm is developed for our estimation problem, which offers benefits in terms of smaller computational load in comparison to standard CKF. Furthermore, a scheme which is termed as adaptive-RBCKF (A-RBCKF) has been designed for adaptation of process noise covariance. A comparative study of the two state estimators, that is, RBCKF and A-RBCKF, through simulations under different transient scenarios demonstrates the efficacy of A-RBCKF over RBCKF.