A Modified Self-adaptive Particle Swarm Optimization Algorithm

Based on the analysis of inertia weight of the standard PSO, a PSO method is described with self-adaptive stochastic inertia weight based on diversity of individual location and fitness value. Position and fitness value correspond to the axis, based on the difference of location and fitness value from the generation and the current generation to construct a right triangle. It is to modify the inertia weight by change of hypotenuse. By the experiments of six functions, compared with standard PSO and algorithm from the literature, experimental result show that the new algorithm cost lower running time and faster convergence, improved the overall performance.