Neutron/Gamma Discrimination Utilizing Fuzzy C-Means Clustering of the Signal from the Liquid Scintillator

The fuzzy c-means (FCM) clustering method was applied to the neutron/gamma discrimination of the pulses from the liquid scintillator. An experimental setup termed the portable real-time n/γ discriminator with a BC-501A liquid scintillator detector was used to collect waveforms with a 500 Ms/s, 12 bit sampling ADC. The FCM clustering and PGA were applied to the same pulses dataset respectively and the results were compared to each other. Compared to the PGA, the FCM clustering decreased the uncertainty thus improved the discrimination performance. The implementation of FCM clustering in the digital devices also reduced the cost and simplified the algorithm.