PSO (Particle Swarm Optimization) for Interpretation of Magnetic Anomalies Caused by Simple Geometrical Structures

Khalid S. Essa,Mahmoud Elhussein

DOI: https://doi.org/10.1007/s00024-018-1867-0

IF: 2.641

2018-04-19

Pure and Applied Geophysics

Abstract:A new efficient approach to estimate parameters that controlled the source dimensions from magnetic anomaly profile data in light of PSO algorithm (particle swarm optimization) has been presented. The PSO algorithm has been connected in interpreting the magnetic anomaly profiles data onto a new formula for isolated sources embedded in the subsurface. The model parameters deciphered here are the depth of the body, the amplitude coefficient, the angle of effective magnetization, the shape factor and the horizontal coordinates of the source. The model parameters evaluated by the present technique, generally the depth of the covered structures were observed to be in astounding concurrence with the real parameters. The root mean square (RMS) error is considered as a criterion in estimating the misfit between the observed and computed anomalies. Inversion of noise-free synthetic data, noisy synthetic data which contains different levels of random noise (5, 10, 15 and 20%) as well as multiple structures and in additional two real-field data from USA and Egypt exhibits the viability of the approach. Thus, the final results of the different parameters are matched with those given in the published literature and from geologic results.

geochemistry & geophysics