Geoacoustic Parameters Inversion by Combining Particle Filter and Genetic Algorithm

The spatial variability of seabed properties often has a strong impact on sound propagation in shallow water. The range-dependent geoacoustic inversion has been a hot topic recently. This article proposes a method that combines genetic algorithm (GA) and particle filtering (PF) to solve range-dependent inversion problem. GA has good global optimality and PF can make use of prior information effectively. The method proposed in this article uses GA to initialize the particles in PF, and then feeds the particles to GA after PF. This method can improve the concentration of particles in PF. The simulation results demonstrate that compared with PF and other methods, GA+PF gives a better estimation of the geoacoustic parameters and the thickness of the sediment layers when the seabed properties change dramatically.