Introduction Muon scattering tomography(MST)can be employed to scan cargo containers and vehicles for special nuclear materials by using cosmic muons.However,the flux of cosmic ray muons is relatively low for direct d...Introduction Muon scattering tomography(MST)can be employed to scan cargo containers and vehicles for special nuclear materials by using cosmic muons.However,the flux of cosmic ray muons is relatively low for direct detection.Thus,the detection has to be done in a short timescale with small numbers of muons to satisfy the demands of practical applications.Method In this paper,we propose an artificial neural network(ANN)algorithm for material discrimination using MST.The muon scattering angles were simulated using Geant4 to formulate the training set,and the muon scatter angles were measured by Micromegas detection system to create the test set.Results The ANN-based algorithm presented here ensures a discrimination accuracy of 98.0%between aluminum,copper and tungsten in a 5 min measurement of 4×4×4 cm^(3)blocks.展开更多
基金supported by the Program of National Natural Science Foundation of China Grant Nos.11805168 and 21805251
文摘Introduction Muon scattering tomography(MST)can be employed to scan cargo containers and vehicles for special nuclear materials by using cosmic muons.However,the flux of cosmic ray muons is relatively low for direct detection.Thus,the detection has to be done in a short timescale with small numbers of muons to satisfy the demands of practical applications.Method In this paper,we propose an artificial neural network(ANN)algorithm for material discrimination using MST.The muon scattering angles were simulated using Geant4 to formulate the training set,and the muon scatter angles were measured by Micromegas detection system to create the test set.Results The ANN-based algorithm presented here ensures a discrimination accuracy of 98.0%between aluminum,copper and tungsten in a 5 min measurement of 4×4×4 cm^(3)blocks.
