Adaptive particle swarm optimization algorithm based on levy flights mechanism

This article aims to the problems that the particle swarm optimization (PSO) algorithm has slow convergence and easy to fall into local optimum, provides an improved adaptive particle swarm optimization algorithm based on Levy flight mechanism (LFAPSO). The long jumps of Levy flight will step out of the local optimum in the local search. The convergence speed and accuracy of the LFAPSO algorithm are certified on 6 typical test functions. The simulation results show that the LFAPSO algorithm is obviously more successful than chaotic particle swarm optimization algorithm with adaptive mutation (ACPSO) and adaptive particle swarm optimization (APSO) algorithm in convergence performance and robustness. Furthermore, the results demonstrate the LFAPSO algorithm works better to solve the multidimensional function. The method will be used to different optimization problems such as scheduling problems, training neural networks, image segmentation, etc.