Asynchronous and stochastic dimension updating PSO and its application to parameter estimation for frequency modulated (FM) sound waves

The particle velocity and position updating play very important roles for achieving good optimization performance of Particle Swarm Optimization (PSO). This paper analyzed the performance of asynchronously updating PSO and synchronously updating PSO by simulation and discovered that the asynchronously updating way can achieve better optimization performance than the synchronously updating way. Moreover, the convergence rate of asynchronously PSO is faster than that of synchronously PSO, which means that there is spare time to achieve better optimization performance based on certain techniques. Here we proposed a stochastic dimension updating technique in which only some dimensions of position will be updated. Several benchmark functions have been used to validate the proposed method. The proposed method is also applied to the parameter estimation for frequency modulated sound waves.