【摘要】 针对基于传统核材料识别系统中特征值易受系统噪声影响且精确度不够的问题,将高阶统计特征概念引入252 Cf源驱动核系统特征提取和识别之中,提出了一种基于互相关函数与高阶统计特征的252 Cf源驱动核材料富集度识别方法。通过互相关运算及高阶统计分析,得到核系统信息的三维特征图像,并在此基础上开展了待测核材料(235 U)的富集度实验研究与分析识别工作。实验研究结果表明,该识别方法能够较好地降低背景辐射噪声与电子学系统噪声。与传统识别方法相比,对于核部件富集度变化,该算法的敏感性与鲁棒性均有大幅提高。更多还原

【Abstract】 In order to enhance the sensitivity of nuclear material identification system(NMIS) against the change of nuclear material enrichment,the principle of high order statistic feature is introduced and applied to traditional NMIS.We present a new enrichment identification method based on cross-correlation and high order spectrum algorithm.By applying the identification method to NMIS,the 3D graphs with nuclear material character are presented and can be used as new signatures to identify the enrichment of nuclear materials.The simulation result shows that the identification method could suppress the background noises,electronic system noises,and improve the sensitivity against enrichment change to exponential order with no system structure modification.更多还原