摘要：Glowworm swarm optimization (GSO) algorithm is a new intelligent optimization algorithm. Based on the problems of GSO, such as easy to fall into local optimum, slow convergence speed and low optimization precision, an improved GSO with group division is presented. Using shuffled frog leaping algorithm (SFLA), glowworms are divide into different ethnic groups, and local search and global information exchange method improves the GSO performance. The mechanism based on particle position update mechanism in PSO is proposed in order to improve glowworm diversity. By using chaos optimization technique, glowworm groups are initialized, and the algorithm can obtain high quality initial solutions group. Finally, with the classical test functions, the simulation results show that, the GSO with hybrid behavior has better convergence speed and precision. According to the different types of firefly and cold light color is not the same, the glowworm swarm is divided into two sub group, to complete the aspects of paired glowworm swarm population quantity change. Then the cloth Valley bird search algorithm, cloth Valley bird by Levi to fly to the best way to choose size, this kind of flying mode with the machine more strong, will this flight mode into two populations of fireflies swarm evolutionary algorithm. Finish the fireflies optimization path of improvement
