基于Walsh-Hadamard变换与栈式自编码快速核素识别

Fast Radionuclide Identification Algorithm Based on Walsh-Hadamard Transform and Stacked Auto-Encoders

为解决短时辐射环境核素组成分析问题,提出了基于Walsh-Hadamard变换与栈式自编码快速核素识别算法。该算法采用探测率和精确核素识别能谱计数两个参数作为评价新型核素识别算法性能的指标。实验结果表明:新核素识别算法在满足精确核素识别能谱计数的条件下,性能不受探测时间、核素个数和探测距离的影响,而且对运动放射源具有较好的识别性能。新核素识别算法可用于辐射检测设备的快速核素识别。

A fast radionuclide identification algorithm based on Walsh-Hadamard transform and Stacked Auto-Encoders was developed in this paper,with the aim of identifying the isotopic compositions of radiation environment in a short time. Detection rate and accurate radionuclide identification spectrum count were used as the index of evaluating the performance of the proposed algorithm in source identification. The results of the present study indicated that the identification performance of the proposed algorithm was not affected by detection time,number of radionuclide,and detection distance when the accurate radionuclide identification spectrum count of a single radionuclide was met. Moreover,the proposed method exhibited a better performance in identifying moving radionuclide. Therefore,the proposed algorithm has the potential application in radiation monitoring instruments for fast radionuclide identification.