摘要
针对一种新的群集智能——自由搜索优化的不足,提出了基于粗细粒交叉的搜索算法.该算法定义了粗粒交叉和细粒交叉两种算子.通过粗粒交叉,有利于产生新的优秀个体,提高算法的全局搜索能力;采用细粒交叉,在搜索半径内产生更多的优良基因,提高局部搜索能力.典型函数的实验结果表明:新算法的收敛速度、收敛精度、鲁棒性和稳定性大大优于基本自由搜索优化和标准微粒群算法.
A novel free search algorithm based on coarse-grained and fine-grained cossover is proposed by combining free search with genetic algorithm. The algorithm defines two basis operators including coarse-grained cossover and fine-grained cossover. The operators of coarse-grained cossover make the algorithm obtain strong global exploring ability, and the operators of fine-grained cossover make the algorithm have strong local searching ability. Experimental results show that the convergence speed, the convergence probablity, the robustness and the stability of the algorithm are better than those of basic free search algorithm and particle swarm optimization.
出处
《控制与决策》
EI
CSCD
北大核心
2008年第9期1068-1072,共5页
Control and Decision
基金
国家自然科学基金项目(60474076)
江苏省高校自然科学基础研究项目(07KJB510095)
关键词
群集智能
自由搜索优化
微粒群算法
交叉
遗传算法
Swarm intelligence
Free search（FS）
Particle swarm optimization（PSO）
Crossover
Genetic algorithm
