Comparison of Performance of Heuristic Search Methods for Phase Velocity Inversion in Shallow Surface Wave Method

Abstract

Three heuristic search methods, Genetic Algorithm, Simulated Annealing and Tabu Search are implemented to invert Rayleigh wave phase velocity for shallow S-wave velocity profiling in seismic surface wave surveying. Unlike linearized least-squares inversion, they do not require derivative calculation, or an initial model, only the forward modeling calculation. In this study, the performances of the three heuristic techniques are compared with numerical experiments. With common paramerization and search limits, the three algorithms can accurately reconstruct shallow S-wave profiles with and without a velocity reversal from the synthetic data without any specific initial models. The Genetic Algorithms and the Simulated Annealing show fast convergence. However, the Simulated Annealing can find models with the smallest misfits. It is also found that the decrease of the misfits is the most gradual in the Tabu Search.

Application of particle swarm optimization to interpret Rayleigh wave dispersion curves

“…Following Shaw and Srivastava (2007), we set the probabilities of selection, crossover, and mutation as 0.8, 0.6, and 0.02, respectively, in this study. Pezeshk and Zarrabi (2005) and Yamanaka (2005) have also used similar tuning parameters to perform a Rayleigh wave dispersion curve inversion scheme by GA. However, when GA are used for importance sampling purposes, mutation probability should be higher than 0.1 (Fernández .…”

“…Consequently, local linearized inversion strategies are prone to being trapped by local minima, and their success depends heavily on the choice of the initial model and on the accuracy of partial derivative calculations (Cercato, 2009). Thus, global optimization methods that can overcome this limitation are particularly attractive for surface wave analysis, such as genetic algorithms Pezeshk and Zarrabi, 2005;Yamanaka, 2005;Yamanaka and Ishida, 1996), simulating annealing (Beaty and Schmitt, 2003;Beaty et al, 2002;Pei et al, 2007;Ryden and Park, 2006), artificial neural network (Shirazi et al, 2009), wavelet transform (Tillmann, 2005), Monte Carlo (Foti et al, 2009;Socco and Boiero, 2008), and pattern search algorithms (Song et al, 2008(Song et al, , 2009). Nature has always been an inspiration source for scientists.…”

System dynamics simulation and optimization with fuzzy logic

“…We proposed the use of GA to optimize model parameters in order to produce the target output curve. Where the non-linear nature of SD models poses problems for other optimization approaches, GA allows for a relatively quick exploration of the solution space in order to arrive at a satisfactory solution [7,8].…”

Determination of elastic properties of shallow sedimentary deposits applying a spatial autocorrelation method

“…These algorithms are very convenient when the researcher lacks previous information about the ground conditions, in spite of larger and time-consuming computations. The capabilities of the Genetic Algorithms have been compared with some other heuristic method by Yamanaka (2005).…”