Study on asynchronous update mechanism in Particle Swarm Optimization

As a result of easy to implement and few parameter to adjust, Particle Swarm Optimization (PSO) have extensively applied in various engineering fields. Most of modifications on PSO currently emphasized on update mechanism of velocity for new populations, few literatures considered asynchronous communication among the swam during iterative processes. This paper proposes asynchronous update strategy based on an improved switching PSO, and then presents a performance study on asynchronous update strategy. The algorithms known as adaptive weight PSO with synchronous and asynchronous progresses are tested on several benchmark functions. Numerical results show advantages of convergence performance and quality of solution, in comparison with algorithms without asynchronous procedure.