Study on a digital pulse processing algorithm based on template-matching for high-throughput spectroscopy

Abstract

a b s t r a c t A major challenge in utilizing spectroscopy techniques for nuclear safeguards is to perform highresolution measurements at an ultra-high throughput rate. Traditionally, piled-up pulses are rejected to ensure good energy resolution. To improve throughput rate, high-pass filters are normally implemented to shorten pulses. However, this reduces signal-to-noise ratio and causes degradation in energy resolution. In this work, a pulse pile-up recovery algorithm based on template-matching was proved to be an effective approach to achieve high-throughput gamma ray spectroscopy. First, a discussion of the algorithm was given in detail. Second, the algorithm was then successfully utilized to process simulated piled-up pulses from a scintillator detector. Third, the algorithm was implemented to analyze high rate data from a NaI detector, a silicon drift detector and a HPGe detector. The promising results demonstrated the capability of this algorithm to achieve high-throughput rate without significant sacrifice in energy resolution. The performance of the template-matching algorithm was also compared with traditional shaping methods.

Digital pulse processing for NaI(Tl) detectors

“…This result is consistent with NaI(Tl) energy resolution estimates in similar irradiation conditions (i.e. Cs-137 and analog acquisition chain), which range from 8.2% for large modules (50.8 cm length, 20.3 cm diameter) [13] to 6.3% for smaller crystals (5.08 cm length, 5.08 cm diameter) [14]. The digital numerical integral approach yielded a 6.63% energy resolution, using 360 pulse samples.…”

Перспективные Цифровые Решения Для Считывающей Электроники Кремниевых Детекторов

“…Для этого создается матрица М откликов детектора от последовательности дельта функций x 0 (t), случайных по времени и амплитуде, то есть в этом случае выходной сигнал представляет собой набор импульсных характеристик. Выходной сигнал y(t) является прямой сверткой: y(t) = x(t) * М. Тогда оценка амплитуды входного сигнала x(t) может быть получена посредством обратной свертки x(t) = y(t) * М -1 (алгоритм the template-matching algorithm) [11].…”