A Rayleigh Wave Globally Optimal Full Waveform Inversion Framework Based on GPU Parallel Computing

Conventional gradient-based full waveform inversion (FWI) is a local optimization, which highly dependent on the initial model and prone to trapping in minima. Globally optimal FWI that can overcome this limitation particularly attractive, but currently limited by huge amount of calculation. In paper, we propose globally framework based GPU parallel computing, greatly improves efficiency, expected make more widely used. framework, simplify recombine parameters, optimize iteratively. Each iteration contains hundreds individuals, each individual independent other, forward modeling cost function The suitable for variety algorithms, test with particle swarm optimization algorithm example. Both synthetic field examples achieve good results, indicating effectiveness framework.