摘要
针对多维背包问题,提出了一种改进的布谷鸟搜索算法(Modified Cuckoo Search Algorithm,MCS)。该算法保留了基本布谷鸟搜索算法在实数域中的莱维飞行特征,只对其进行截断取整操作,采用了异或操作将鸟蛋被主人发现后位置的随机生成定义到0-1空间。通过对典型多维背包问题的仿真实验和与基本布谷鸟搜索算法、二进制微粒群算法和禁忌搜索算法的比较,表明了所提出的算法的收敛速度更快,全局寻优能力更强。
A modified cuckoo search algorithm (MCS) is proposed for solving multi-dimensional knapsack problems. In MCS, the 16vy flight characteristics of the basic cuckoo search algorithm(CS) in the real domain is retained, and just only with tnmcation integer operation, xor operation is used to define the method to generate a new position in the 0-1 space randomly when the eggs are found by the host. Simulations of the multi-dimensional knapsack problem and comparisons with CS and the binary particle swarm optimization algorithm (BPSO) and tabu search algorithm (TS) show that the convergence speed of MCS is faster, and the global optimization ability is stronger.
出处
《控制工程》
CSCD
北大核心
2016年第7期1069-1075,共7页
Control Engineering of China
基金
国家自然科学基金资助项目(71401106)
高等学校博士学科点专项科研基金联合资助课题(20123120120005)
上海市教育委员会科研创新项目(14YZ090)
关键词
多维背包问题
布谷鸟搜索算法
优化
Multi-dimensional knapsack problem
cuckoo search algorithm
optimization