#TITLE_ALTERNATIVE#

• 主要作者: MAULANA (NIM12306009) pembimbing : Dr. Hendra Grandis, YAHYA
• 格式: Final Project
• 語言: Indonesia
• 在線閱讀: https://digilib.itb.ac.id/gdl/view/17600
• 機構: Institut Teknologi Bandung

One-dimensional magnetotelluric inversion modeling is useful for estimating the distribution of subsurface physical parameters i.e. resistivity which varies with depth. A non-linear relationship between model parameters with observation data results in difficulties in the inversion process. This problem can be solved using non-linear inversion method with a global approach. Particle Swarm Optimization (PSO) is an optimization method based on the behavior patterns of a group of animals in achieving common goal. This optimization method is use to solve the non-linear inversion with global approach. In this concept, particles are represented as models. In PSO algorithm, position of each particle (individual) originally at random will be modified iteratively by its best position of individual (cognitive learning term) and the group’s best position (social learning term). The best positions are associated with minimum misfit value during iteration process. This algorithm was applied to invert one- dimensional magnetotelluric synthetic data. The synthetic data are as responses of simple three-layer synthetic models. Misfit is the sum of quadratic difference between observed data and model response (calculated data). Tests conducted using different numbers of particles with the same iteration number didn’t result in significantly different. The optimal results were obtained by 200 particles with 400 iterations. This result show the PSO algorithm can be applied to the inversion modeling of one-dimensional magnetotelluric data.