neighborhood algorithm

This study presents a novel approach for inverting surface wave dispersion curves using the Ant Colony Optimization (ACO) algorithm in the MATLAB environment. Due to the problem's nonlinear nature and multiple extrema, this method demonstrates significant advantages over local search techniques. To evaluate the performance of the proposed algorithm, synthetic data from three six-layer geological models were analyzed. In these models, the velocity and thickness values were defined over a wide range, and the velocity variation with depth was determined based on the λ/2 rule. The simultaneous inversion process involved estimating shear wave velocity, compressional wave velocity, and layer thickness, assuming a variable Poisson’s ratio (0.1 to 0.5) and a constant density. The results indicate that the average relative error in velocity parameters using the ACO method is 3.5%, whereas this error reaches 15% for the neighborhood algorithm available in the Geopsy software. These findings highlight the superiority of the proposed approach in terms of computational efficiency and reliability. Additional advantages include its independence from an initial model, the ability to explore the entire parameter space, and reduced sensitivity to local extrema. Finally, the algorithm was applied to experimental field data from a station in the Tabriz Plain, yielding results that closely matched available borehole data. Based on these findings, a 2D shear wave velocity model for the region was developed.