Pulsed Fast Neutron Analysis Effective Mine-detection Signal Acquisition And Resolution

Thousands of new victims are caused by the remained landmines each year estimated by the ICBL. Numerous demining methods have been developed to change this situation, among which the nuclear techniques showes a number of advantages for non-destructive elemental characterization. Pulsed fast/thermal neutron analysis (PFTNA) is an advanced neutron-based technique, which has the capability to quantify the elemental contents and the ratios of the target object. In some developed countries headed by the United States, the PFTNA method has been employed in the humanitarian demining area. PELAN is a portable device developed by the Western Kentucky University, which has successfully undergone field demonstrations for landmine detection. However, the domestic researches on PFTNA have just developed into the preliminary stage. This paper is the first attempt to identify the landmines ultilize the PFTNA method.The detailed processes of the source neutron from birth to death were simulated through the Monte Carlo method. The time-scale of each process was provided and the principle of PFTNA was verified by the simulation results. Then, based on the criteria of "Most Benificical to the Thermal Spectrum", an innovative theorectical model was proposed to acquire the optimal PFTNA timing parameters. Feasibility and reliability of this model were evaluated under different application scenarios.The key components of the PFTNA demining system such as moderator and shield system were designed through MCNP5. Then the overall detection layout of the experimental system was optimized and accomplished. The PFTNA time-share circuit was successfully developed and based on which the landmine detection experiment was carried out. Thermal neutron induced spectrums and fast neutron induced spectrums of typical target objects were collected successfully. The preliminary analysis result showed that the detection system can distinguish landmines from common distracters, so the feasibility of the PFTNA demining method was demonstrated. But in terms of detection speed, the existing system still can not meet the actual needs of the landmine detection.A comparison was drawn between two spectroscopy analysis algorithms:the traditional Least-Square algorithm and the new principal component algorithm. The PCA algorithm was probed theoretically to analyse PFTNA spectrums. A simplified PFTNA detection model was constructed through the Geant4program and the spectrums of nine known samples were applied to train this model. Through prophase learning, the system possessed the capability to distinguish explosives from ordinary organic objets. Then the feasibility of PCA algorithm was proved by this process.Both hardware and software deficiencies of exsisting PFTNA landmine detection system were assessed. Then a detailed plan was made to improve the performance of the system. Finally, an outlook was made upon PFTNA method applying to the future humanitarian demining career.