Abstract:

We divide underground medium into a certain number of grid cells and construct an exponential function that simulates the magnetic distribution at the center of the magnetic body grid, and through Particle Swarm Optimization(PSO) algorithm, the functional parameters is inversed. Then, according to the parameter of the function, the magnetization of each grid cell is computed, from which, magnetization fast imaging is carried out. Theoretical model test shows that this method not only is capable of solving a physical property distribution of a 20 × 10 grid-scale problem simplifying to a problem of solving 10 parameters of an exponential function, but also, when adding 20% random noises to anomalies of different geological models, can achieves satisfactory result with fewer adjusted parameters. Thus demonstrates the robust of the algorithm. Applying this method to the process of a field magnetic profile over the iron ore district in Palishan, Laos, a strong magnetic body with a thickness of nearly 140 m was obtained, confirmed by the drilling which reveals the cumulative maximum thickness 130.6 m of the actual orebody. This further indicates the feasibility and effectiveness of this method for practical data processing and interpretation.

Key words: particle swarm optimization(PSO), magnetization imaging, magnetic distribution model, Palishan in Laos, magnetic fields