The Age And Nuclear Abundance Quantitative Identification Of Uranium By RBF Network

Nuclear material identification technology has been used widely in nuclear weapons examination,monitoring nuclear materials,preventing nuclear from difusing and prevention of nuclear terrorism.Age and nuclear abundance are important characteristic parameters?feature quantity?of characteristic nuclear materials,it is also an important basis of identifying the types of nuclear materials or the origin and the history of nuclear materials,Therefore,it is necessary to quantitatively identify the age and abundance of nuclear materials.Radionuclide in nuclear materials is accompanied by characteristic?-rays in its decaying process,and because?-rays created has its own uniqueness,the?-ray energy in the nuclear materials can be used as a"radiation fingerprint"to sign and identify nuclear materials.The purpose of this work is to regard?spectrum fingerprint as identification features,and then establish the mapping relationship between?spectrum fingerprint and nuclear material age or nuclear abundance with the RBF artificial neural network,and finally achieve the quantitative identification of the Uranium nuclear material age and the nuclear abundance through pattern recognition.In this work,Uranium nuclear materials were studied.At first,the?spectrum fingerprints of the known Uranium nuclear materials were simulated with Monte Carlo,then the?spectrum fingerprints of known Uranium nuclear materials are used as input and relevant age or nuclide abundance as output to train the RBF artificial neural network,Finally unknown samples are used to verify the artificial neural network,The main contents of this work are as follows:?1?Quantitative identification of Uranium nuclear material age.By establishing mapping relationships between?spectrum fingerprints and the ages of Uranium nuclear materials,?spectrum fingerprints were identified by single group and multiple groups separately.?2?Identification ability testing of the networkThe identification ability of the network was verified by identifying the ages of unknown Uranium nuclear materials besides the training samples.?3?Quantitative identification of nuclear abundance of Uranium nuclear materialsBy establishing a mapping relationship between the?spectrum fingerprints and the nuclide abundances of uranium nuclear materials,and establishing training sample set with random sampling method,quantitative identification of the nuclide abundances in unknown uranium nuclear materials were achieved by RBF network.The research shows that the relative identification errors of the age about single group and multiple groups of uranium nuclear materials are less than 7%;Based on sample U4 to change,nuclide abundance of ²³⁵U varies within the range of-8%to+5%,then the relative recognition error of the age is within 10%.Based on sample U5 to change,nuclide abundance of²³⁵U varies within the range of-14%to+7%,then the relative recognition error of the age is within 10%.And it can be concluded that as the magnitude of abundance becomes increasing,the relative error of recognition increases as well;The relative error for the identification of nuclide abundance?current nuclear abundance?of Uranium nuclear materials is less than 7.37%;In conclusion,it is effective and feasible to identify the ages and nuclear abundances of Uranium nuclear materials quantitatively by?spectrum fingerprint recognition technology through the RBF artificial neural network.