Differential evolution algorithm for nonlinear inversion of high-frequency Rayleigh wave dispersion curves

摘要: Abstract In recent years, Rayleigh waves are gaining popularity to obtain near-surface shear (S)-wave velocity profiles. However, inversion of wave dispersion curves is challenging for most local-search methods due its high nonlinearity and multimodality. this study, we proposed tested a new curve scheme based on differential evolution (DE) algorithm. DE novel stochastic search approach that possesses several attractive advantages: (1) Capable handling non-differentiable, non-linear multimodal objective functions because strategy; (2) Parallelizability cope with computation intensive without being time consuming by using vector population where the perturbation vectors can be done independently; (3) Ease use, i.e. few control variables steer minimization/maximization DE's self-organizing scheme; (4) Good convergence properties. The inverse procedure was applied nonlinear fundamental-mode S-wave To evaluate calculation efficiency stability DE, firstly inverted four noise-free noisy synthetic data sets. Secondly, investigated effects number layers algorithm made an uncertainty appraisal analysis Thirdly, comparative genetic algorithms (GA) set further investigate performance procedure. Finally, real-world example from waste disposal site in NE Italy examine applicability curves. Furthermore, compared GA scores described here. Results both actual field demonstrate high-frequency surface should considered good not only terms accuracy but also speed.