Improved Particle Swarm Optimization With Dynamically Changing Inertia Weight

摘要

In order to improve the performance of particle swarm (PSO) algorithm which inertia weight was decreased linearly, a novel particle swarm optimization (NPSO) algorithm with dynamically changing inertia weight was presented. In each iteration process, the inertia weight of the improved algorithm was changed dynamically based on the current iteration and the best fitness. The new algorithm was tested with three benchmark functions. The test results indicated that the disadvantages of slow speed on convergence and easy to be trapped in local optimum of the linearly decreasing weight of the PSO could be overcome effectively.