Nuclear counting filter based on a Centered Skellam Test and a double Exponential Smoothing

R. Coulon (1) , Vladimir Kondrasovs (1) , J. Dumazert (1) , E. Rohée (1) , S. Normand (1)

1 LCAE (CEA, LIST) - Laboratoire Capteurs et Architectures Electroniques (CEA, LIST)

Online nuclear counting represents a challenge due to the stochastic nature of radioactivity. The count data have to be filtered in order to provide a precise and accurate estimation of the count rate, this with a response time compatible with the application in view. An innovative filter is presented in this paper addressing this issue. It is a nonlinear filter based on a Centered Skellam Test (CST) giving a local maximum likelihood estimation of the signal based on a Poisson distribution assumption. This nonlinear approach allows to smooth the counting signal while maintaining a fast response when brutal change activity occur. The filter has been improved by the implementation of a Brown's double Exponential Smoothing (BES). The filter has been validated and compared to other state of the art smoothing filters. The CST-BES filter shows a significant improvement compared to all tested smoothing filters.

Mots clés: Filter; Signal Processing; Nuclear; Measurement; nuclear counting filter; Poisson distribution; radioactivity; radioactivity measurement; local maximum likelihood estimation; nonlinear filter; stochastic nature; Brown double exponential smoothing; centered Skellam test; maximum likelihood estimation; Index Terms-Nuclear

Domaines: Physique Nucléaire Expérimentale [nucl-ex]; Traitement du signal et de l'image [eess.SP]; Physique Numérique [physics.comp-ph]; Méthodologie [stat.ME]

R. Coulon, Vladimir Kondrasovs, J. Dumazert, E. Rohée, S. Normand. Nuclear counting filter based on a Centered Skellam Test and a double Exponential Smoothing. 2015 4th International Conference on Advancements in Nuclear Instrumentation Measurement Methods and their Applications (ANIMMA), Apr 2015, Lisbon, Portugal. pp.7465500, ⟨10.1109/ANIMMA.2015.7465500⟩. ⟨cea-01823368⟩