Researches On Fireworks Explosion Optimization Algorithm And Its Improvement

As a novel kind of optimization approaches developed in recent years, evolutionary algorithms have less limitation about the constrained conditions and goal functions of the optimized problems to be solved compared with the traditional methods. The ideas of the evolutionary algorithms are mostly based on the imitation of some natural phenomena, so they are characterized with artificial intelligence. Although the evolutionary algorithms have been successfully applied in different areas, they are"problem depended"essentially and there is no omnipotent method. So, seeking for a highly efficient algorithm is still a valuable task, which has become a hot topic in optimization researching area.Recently,a novel fireworks explosion optimization (FEO) algorithm based on the idea of fireworks explosion is introduced. The proposed algorithm generates a certain number of fireworks bombs in the search space and each fireworks bomb will implement the operation of explosion, which can generate plenty of sparks to explore in the neighborhood of the original fireworks bomb. The proposed method has been tested on several benchmark functions. Primary results have indicated the good performance of this new algorithm in terms of convergence rate, precision and stabilization.Firstly, the mainly differences between evolutionary approaches and traditional methods are discussed, and the basic ideas and procedures of some classical algorithms are introduced subsequently. The major research interests and main achievements have also been discussed. Then, based on the general structures of the evolutionary approaches, a new algorithm named Fireworks Explosion Optimization was proposed, at the same time, the idea, mathematical model and steps of this algorithm are described in detail. The good performance of this new algorithm is also indicated with experiments on stander test functions.To solve the problem encountered by the FEO algorithm, the paper has proposed the concepts of optimal direction and efficient explosion space and developed a improved algorithm named IFEO, which contains new explosion mechanism and high-precision local search method. Extensive experiments are designed for further investigating on the efficiency of this improved method.