改进的蚁群算法及其在南岭地区花岗岩侵入体探测中的应用  

Improved Ant Colony Optimization and Its Application in Prospecting for Granite Intrusion in Nanling Region

作　　者：刘双[1] 陈国雄[2] 刘天佑[1] 

机构地区：[1]中国地质大学地球物理与空间信息学院,武汉430074 [2]中国地质大学地质过程与矿产国家重点实验室,武汉430074

出　　处：《吉林大学学报（地球科学版）》2013年第6期2005-2017,共13页Journal of Jilin University：Earth Science Edition

基　　金：中国地质调查局地质大调查项目(1212011121101)

摘　　要：蚁群算法(ACO)主要应用于组合优化问题(COPs),很少应用于地球物理重磁资料物性反演。首先,对蚁群系统中目标函数值与信息素的映射机制进行改进。理论模拟表明,改进后的Gauss模型在收敛速度和反演精度方面比传统的ant-cycle模型要好,适用于重磁资料物性反演。然后,将蚁群算法应用于南岭地区千里山岩体和九嶷山岩体重力资料反演与解释。结果表明:蚁群算法圈定的花岗岩岩体与地面地质填图结果及大地电磁测深反演结果吻合,较好地控制了千里山岩体和九嶷山岩体的深度和形状。该结果可为研究南岭地区花岗岩侵入模式和多金属成矿规律提供参考。Ant colony optimization （ACO） is mainly applied to solve combinatorial optimization problems （COPs）, but not yet used to recover the property distribution from gravity and magnetic data. We improved the mapping mechanism from objective function to pheromone amount of ants system. Theoretical simulation showes that the improved Gauss system of ACO is superior to traditional ant- cycle system regarding to convergence speed and fittingness resolution, thus the new method is more suitable to potential field data inversion. The method was applied to the gravity anomalies passing through Qianlishan rockmass and Jiuyishan rockmass in Nangling region, respectively. The granite rocks delineated by ACO inversion are in good agreement with surface geology and two-dimensional magnetotelluric （MT） inversion model. The results provide references of granite intrusion pattern and metallogenic regularity in Nanling region.

关 键 词：蚁群算法 ant—cycle模型 GAUSS模型 南岭花岗岩侵入体 重力异常 千里山岩体 九嶷山岩体 

分 类 号：P631[天文地球—地质矿产勘探]