Nuclear Science and Engineering / Volume 160 / Number 3 / November 2008 / Pages 302-317
A wavelet ridge application is proposed as a simple method to determine the evolution of the linear stability parameters of a boiling water reactor nuclear power plant (NPP) using neutronic noise signals. The wavelet ridges are used to track the instantaneous frequencies contained in a signal and to estimate the decay ratio (DR). The first step of the method consists of denoising the analyzed signals by a discrete wavelet transform to reduce the interference of high-frequency noise and concentrate the analysis in the band where crucial frequencies are presented. Next is computation of the wavelet ridges by a continuous wavelet transform to obtain the modulus maxima from the normalized scalogram of the signal. In general, associations with these wavelet ridges can be used to compute the instantaneous frequency contained in the signal and the DR evolution with the measurement. To study the performance of the wavelet ridge method, by computing the evolution of the linear stability parameters, both simulated and real neutronic signals were considered. The simulated signal is used to validate methodically and to study some features of the wavelet ridge method. To demonstrate the method applicability, three real neutronic signals related to instability events in the Laguna Verde NPP and Ringhals and Forsmark stability benchmarks were analyzed. The investigations show that most of the local energies of the signal are concentrated and that DR variations of the signals were observed along the measurements.