Fireworks Explosion Optimization Improvement And Researches On Its Performance Test

In recent years, swarm intelligent optimization algorithms are very popular in the fieldof intelligent optimization, and they have become a new kind of evolutionary computationtechniques. Compared with traditional optimization methods, swarm intelligenceoptimization algorithms have better stability and efficiency in solving optimizationproblems, and the algorithms are simple, implicit parallelism, easy to understand and easyto implement. These advantages effectively promoted the application of optimizationtechnology. Thus, the research of swarm intelligence optimization algorithm has somepractical significance.Firstly, this paper introduces the swarm intelligence optimization algorithm andpresent status. Secondly, This paper introduces a novel fireworks explosion optimization(FEO) based on the idea of fireworks explosion, it provides a new method for optimizingcontinuous nonlinear functions. The algorithm implements diffuse parallel search in thesearch space by imitating the process of fireworks explosion. Then, the paper present anovel improved algorithm inspired by the idea of Genetic algorithm, entitled Fireworksexplosion optimization with GA operators (GAFEO). GAFEO algorithm mainlyimplements diffuse parallel search in the search space by imitating the process offireworks explosion, namely, neighborhood of fireworks bomb is diffusely covered withsparks. Then, in order to improve the optimal performance, the algorithm introduces theadaptive local search strategy, crossover and mutation strategies, and burst pointmanagement strategy. Experiments are conducted on 30 benchmark problems whichinclude unimodal and multimodal functions. The results show that the explosion algorithmdisplays better feasibility and effectiveness.