Direct Inversion of Shear-Wave Velocity using Particle Swarm Optimisation Algorithm

The shear-wave velocity is one of the most important parameters for the evaluation of site characteristics. The spatial autocorrelation method is one of several methods developed to obtain the velocity of shear waves via surface wave exploration. However, the accuracy of the inverted shear-wave velocity is evidently affected by manual selection of the dispersion curve in the spatial autocorrelation method. In this paper, we propose a scheme that helps realise the direct inversion of the spatial autocorrelation coefficient (DISPAC) based on particle swarm optimisation (PSO) to avoid picking errors. The particle swarm optimisation seeks the optimal solution using a given parameter through continuous individual optimisation and a swarm optimisation guidance algorithm. Therefore, DISPAC_PSO has the advantage of overcoming the existing uncertainty of fitting a Bessel function to the spatial autocorrelation coefficient. In contrast to common nonlinear algorithms such as simulated annealing and genetic algorithms, particle swarm optimisation has a fast convergence speed and high inversion accuracy. To evaluate the stability and accuracy of the proposed direct inversion of the spatial autocorrelation coefficient with particle swarm optimisation, the proposed method was compared with conventional inversion of the dispersion curve, the results verified that the proposed approach has a relatively high accuracy.