基于RBF神经网络的钚年龄定量识别

Quantitative identification of plutonium age based on RBF neural network

利用RBF网络建立了γ指纹与钚年龄和核素初始丰度之间的映射,以γ指纹为特征,采用模式识别方法实现了钚年龄的定量识别。对已知和未知钚材料γ指纹的识别研究结果表明,基于RBF神经网络的钚年龄模式识别方法充分利用了γ指纹的全部信息,克服了传统γ能谱分析方法的不足,通过建立相应的训练样本集,方法可准确识别未知钚材料的年龄,并可对其历史和来源做出准确推断。

The mapping between y fingerprint and plutonium age,initial nuclide abundances was es-tablished by using RBF neural network. With Y fingerprint as the recognition feather,pattern recogni-tion method was applied to realize quantitative identification of plutonium age. The research shows that by establishing corresponding training sample set,the plutonium age can be accurately identified,and the history and source of the unknown plutonium materials can be accurately deduced.