Comparison among the inversion results of surface wave dispersion curves by the artificial neural networks and local and global search methods

Abstract

We propose a new method for the inversion of surface wave dispersion curves based on the application of an artificial neural network and we suggest a data–driven approach for selecting the range of the space parameters for the calculating training data set. The synthetic data processing results showed that the accuracy of the proposed method is superior local search and equivalent to global search methods, whereas the proposed method is more robust in the presence of noise.