Abstract:

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.

Key words: ant colony optimization (ACO), ant-cycle system, Gauss system, Nanling granite intrusion, gravity anomalies, Qianlishan rocks, Jiuyishan rocks