Simultaneously Optimizing Inertia Weight and Acceleration Coefficients via Introducing New Functions into PSO Algorithm

Abstract Because of the drawbacks of easy premature in initial iteration stages, the low convergence accuracy and slowed-down converging speed in final stages of the particle swarm optimization (PSO)algorithm therefore the simple particle swarm optimization (SPSO) algorithm with dynamic changes of inertia weight and acceleration coefficients (IASPSO) has been put forward. IASPSO algorithm provides a parameter optimization strategy by using exponential decreasing inertia weight and sine function acceleration coefficient to improve global exploration capacity. Simulation tests are carried out with classic Benchmark test functions. The simulation results show that compared with other PSO algorithms, IASPSO algorithm can converge to the better global optimization with a fast converging velocity and high convergence precision, promoting the optimization performance of the algorithm.