Study On The Improvement Of The Artificial Glowworm Swarm Algorithm

Artificial Glowworm Swarm Optimization(GSO) is inspired by the glowworms in nature. The glowworms in nature forage and attract mates in the way of giving lights. The GSO is a swarm intelligent algorithm rising recently. The appearance of the GSO has attracted much attention. The GSO has much strengths comparing to the other swarm intelligent algorithm. GSO has less adjustable parameters and most of the parameters can be a stable value. Besides, the GSO also has some other advantages such as: the less memory needed, the fast computing speed and so on. The GSO has some remarkable advantages, but it still has some problems needs to be settled, such as: easy to fall into local optimum, slow convergence rate and so on. In order to solve these problems, we propose some optimal schemes to improve the algorithm. This paper do the following works and researches:This paper has summarized the research background and significance of GSO, and then introduce the structure of the paper.The paper introduce the GSO in detail, particularly in the theory of the algorithm and the realization. Furthermore, it analyzes the parameters in the algorithm and then summarizes some conclusions related.Propose a scheme to improve the GSO. This paper propose to introduce the bunching and tailgate of Artificial Fish Swarm Algorithm so that it can solve some problems in GSO, such as the GSO is easy to fall into local optimum and its convergence rate is too slow. The bunching introduced chooses the center of the glowworms satisfy the conditions rather than choose randomly. The tailgate adds the congestion factor to GSO, the congestion will decrease the collision during the movement of glowworms. The simulation shows that the GSO improved reduce the iterations effectively.In order to reduce the iterations and the population size at the same time, the paper proposed another improvement. Introduce the Shuffled Leap-Frog Algorithm(SLFA) and Simulated Annealing(SA) to GSO. The group idea of SLFA can reduce the iterations of the GSO and the reception of inferior solution with probabilities of SA can improve the GSO. The simulation shows that the improved algorithm can not only reduce the iterations but also the population size.At the last of the paper, it summarizes the main work and gives some advises of the further work.