一种基于模式识别的放射性核素快速识别方法研究

Fast Radioactive Nuclide Recognition Method Study Based on Pattern Recognition

基于模式识别的方法,应用核辐射数字化测量分析平台,通过综合利用放射性核素的射线信息,选定核素的辐射特征信息,建立核素特征数组库,设计识别方法应用于中、低分辨率探测器的放射性核素识别。在实验中发现在传统中低分辨率谱仪系统对单一核素全能峰未达到统计下限值时,使用快速识别算法的数字系统对3种混合核素的识别率达到90%以上。结果表明此方法明显优于传统算法,有效地提高了放射性核素的快速识别能力。

Based on pattern recognition method , applied the nuclear radiation digital measurement and analysis system platform , through synthetically making use of the radioactive nuclide＇s ray information , selected radiation characteristic information of the radioactive nuclide , established the characteristic arrays database of radioactive nuclides , a recognition method is designed and applied to the identification of radionuclide radiation while using the low-middle resolution detector in the paper .Verified by experiments , when the count value of the tradi-tional low-resolution spectrometer system does not reach single full energy peak ＇s statistical lower limit , true discrimination rate of the three kinds of mixed radioactive nuclides＇reached more than 90%in the digital meas-urement and analysis system using fast radionuclide recognition method .The results show that this fast radioac-tive nuclide recognition method is obviously superior to the traditional methods , and effectively improve the rap-id identification ability to radioactive nuclide .