A study of 2-D magnetotelluric quantum genetic inversion algorithm based on subspace
Quantum Genetic Algorithm is an excellent method. Nevertheless, there is a disadvantage to trap into the local minima for the conventional Quantum Genetic Algorithm. To advance the algorithm and probe the feasibility and effectiveness of the algorithm introduced into the magnetotelluric data 2D inversion, some improvements are made, whose effectiveness is testified through the inversion for ID magnetotelluric two-layer (D-type) model and four-layer (HK-type) model. Then, the improved method is introduced into the magnetotelluric data 2D inversion. Based on the sliding subspace and the most simplified inversion condition, one typical 2D low-resistivity model is inversed using the conventional Quantum Genetic Algorithm and the improved Quantum Genetic Algorithm, respectively. The results indicate that it is feasible and effective to apply the Quantum Genetic Algorithm to magnetotelluric 2D inversion based on the subspace method and the result from the improved method is better than the conventional method. Finally, the better result is also obtained for the field data.
NETL
NSTL
万方数据
维普
