Research On The N/γ Discrimination Method Based On Fuzzy Cluster Analysis

Neutron detection technology has been widely used in the fields of explosives detection, environmental radiation detection, military and deep space exploration. Since almost all neutron fields coexist with an associatedγ-ray component while neutron detectors are also sensitive toγ-ray photons, the discrimination between neutrons and gamma-rays (n/γ) becomes a key technical problem in the field of neutron detection. This paper has focused on this problem and proposed an novel method for the n/γdiscrimination based on Fuzzy C-Means (FCM) clustering.The validity and feasibility of the FCM n/γdiscrimination method was proved by combination of simulation and experimental verification. To begin with, the model of particle pulse signal was built according to the physical character of BC-501A liquid scintillator and the performance of FCM method was researched under different conditions of signal-to-noise ratio. The simulation results showed that the discrimination error ratio of FCM method could be kept under 1% if only the signal-to-noise was smaller than 12dB. Then a neutron detection platform which comprises a BC-501A liquid scintillator detector, a photomultiplier tube R329-02, a high voltage power supply unit C9619-01 and a digital storage oscilloscope DSO6032A was set up to acquire the neutrons and gamma rays in the environment, after which the particle pulse signals were preprocessed by an 25 points moving average filter and meanwhile the the optimal fuzzy exponent m was fixed on. Further more, the FCM clustering and the current method—Pulse Gradient Analysis(PGA) were applied to the same pulses dataset respectively and the results were compared to each other. It is shown that the discrimination results of the FCM clustering were almost consistent with those of PGA when the peaks of the pulse signals were bigger than 1000. But for the pulse signals whose peaks were smaller than 1000, the FCM clustering exhibited better discrimination performance than that of PGA. Ultimately, the cluster centers of neutrons and gamma-rays were found based on abundant tests and the real-time n/γdiscrimination were available.The n/γdiscrimination method based on FCM clustering exhibited a strong insensitivity to noise and anti-interference ability, with which it would has a wide range of applications in the fields of neutron detection, neutron spectrum unfolding, and so on.