采用ANFIS降低GR-660车载γ能谱测量数据的统计噪声  

Reducing statistical noise in GR-660 car gamma-ray spectral data through ANFIS

作　　者：李必红[1] 陆士立[1] 韩绍阳[1] 卫三元[1] 

机构地区：[1]核工业北京地质研究院,北京100029

出　　处：《铀矿地质》2008年第3期170-174,169,共6页Uranium Geology

摘　　要：统计噪声是影响弱信息提取的重要因素。基于MATLAB语言编制的自适应神经网络模糊系统（ANFIS）程序，对2006年在航放标准模型上的车载γ能谱测量标定数据和满洲里地区的车载γ能谱测量数据进行处理的结果表明，该方法极大地减小了标准偏差，相当于探测器体积增加22～1151倍，增强了γ能谱的弱信息，有效地提高了探测精度。A procedure has been developed for reducing statistical noise in GR-660 car gamma-ray spectral data, the result after using ANFIS has shown that reduction in statistic noise is equivalent to increasing the detector volume by a factor typically between 22 and 1151 times. Profiles and maps have been found that the method can significantly enhance weak geological information and effectively improve detecting precision.

关 键 词：自适应神经网络模糊系统 统计噪声γ能谱 

分 类 号：P631.622[天文地球—地质矿产勘探]