Research on Particle Swarm Optimization Based on Dynamic Acceleration Coefficients

For fixed acceleration factors in the particle swarm optimization cause the function optimization accuracy poorly, easy to fall into local optimal solution and slow late convergence, this paper presents an improved particle swarm optimization base on dynamic acceleration coefficients(PSO-DAC).Adopting decreasing inertia weight coefficients improve the ability of weigh local search and global search capability.At the same time, introducing dynamic acceleration coefficients raise the convergence speed and accuracy of the particle swarm algorithm.By four commonly used benchmark functions, the improved particle swarm algorithm contrasts with the standard particle swarm optimization through simulation experiments.The experimental results show, the improved algorithm comparing with the standard particle swarm optimization has higher accuracy and reduce the number of iterations over 51.28%.The optimal solution can be found more quickly, especially in multimodal function.