摘要： 受烟花爆炸现象的启发并结合遗传算法思想提出一种新的优化算法——带有遗传算子的烟花爆炸优化算法(GAFEO)。该算法主要模拟烟花爆炸的方式对解空间进行基本的并行弥漫式爆炸搜索，引入自适应局部搜索策略和遗传算法中的交叉变异策略以改善算法的优化性能。通过实验对12个常用高维测试函数进行优化计算，结果表明，与PSO算法以及其他新型算法相比，GAFEO算法在寻优能力、寻优精度等方面都具有较好的性能。

关键词: 烟花爆炸, 并行搜索, 炸点管理, 交叉变异, 自适应局部搜索

Abstract: This paper presents a new evolutionary algorithm——entitled Fireworks Explosion Optimization with Genetic operators(GAFEO), that is inspired by fireworks explosion and compromises the idea of genetic algorithm. GAFEO algorithm mainly implements diffuse parallel search in the search space by imitating the process of fireworks explosion. In order to improve the optimal performance, the algorithm introduces the adaptive local search strategy and crossover mutation strategies. Experiments are conducted on 12 benchmark problems which include unimodal and multimodal functions. Results show that the GAFEO algorithm displays better performance compared to PSO and other hybrid algorithms.

Key words: fireworks explosion, parallel search, burst point management, crossover mutation, adaptive local search

中图分类号: 

• TP311.52