基于预优共轭梯度的磁化强度反演方法及应用  

The Magnetization Inversion Method Based on Preconditioned Conjugate Gradient Algorithm and its Application

作　　者：曾琴琴[1] 刘双[2] 王永华[1] 

机构地区：[1]成都地质矿产研究所,四川成都610081 [2]中国地质大学地球物理与空间信息学院,湖北武汉430074

出　　处：《地质与勘探》2014年第4期756-762,共7页Geology and Exploration

基　　金：中国地质调查局地质调查工作项目(编号1212011120918;1212011220249和12120113051500)联合资助

摘　　要：基于预优共轭梯度算法的反演方法在求解磁化强度大、规模欠定线性方程组时,可改善方程组的条件数,并提高算法的纵向分辨率。该方法对磁性体形状无要求,且收敛速度快、稳定性好,能较直观地反映磁性体特征。本文通过理论模型试验结果分析证明了基于预优共轭梯度的磁化强度反演方法的稳定可靠性,并对云南北衙铁金矿区万硐山矿段L32线磁异常进行了反演,结果表明了该方法应用于实际资料处理与解释的有效性和可靠性。The inversion method based on preconditioned conjugate gradient algorithm is used to solve large-scale underdetermined linear magnetization equations,which can improve the condition number of equations and vertical resolution of the algorithm. This method not only has no requirements on magnetic body shape,but also has a fast convergence speed and good stability,with which the characteristics of magnetic bodies can be reflected more intuitively. The tests on theoretical models prove this method is stable,reliable and fast. And its use to and the inversion of real magnetic data of exploration line L32 in the Wandongshan ore section of the Beiya iron-gold deposit also indicates the effectiveness and reliability of this method.

关 键 词：预优共轭梯度 深度加权预优因子 磁化强度反演 北衙铁金矿区 

分 类 号：P631.2[天文地球—地质矿产勘探]