Reducción de ruido en el cálculo de la reactividad con filtro digital y cuadraturas de Gauss-Legendre

Abstract

This study presents a method that applies the Gauss-Legendre quadrature to approximate the integral contained in the inverse equation of point kinetics for the calculation of nuclear reactivity. Fluctuations in a neutron population of a nuclear reactor are assumed to be multiplicative Gaussian white noise around the mean. A modified low-pass filter performs fluctuation reductions by considering a symmetrical sampling window for both low and high frequencies. Filter parameters such as the filter constant and the sampling window are optimized for different values of standard deviation. An error surface is estimated from differences between the analytical values and the numerical values of reactivity, which is minimized to determine optimal parameters. It is concluded that the method presented here can be considered as an alternative method to reduce fluctuations in the calculation of nuclear reactivity and can be implemented in a digital reactivity meter.