摘要
针对一维下料优化问题,提出了基于蜂群遗传算法的优化求解方案。具体做法是,以实数表示的各零件长度的一个排列作为一个染色体,其中每个零件的长度作为基因;根据自然界蜂群生物学原理设置了两个种群,一个种群主要用于全局搜索,另一个种群主要用于局部搜索;采用最优个体交叉策略;遗传算子包括联赛选择算子,顺序交叉算子,2-交换变异算子和抑制算子。仿真实验结果表明,该算法逼近理论最优值,而且收敛速度快,较好地解决了一维下料问题。
Presents bee swarm genetic algorithm for one-dimensional cutting stock problem.The concrete means is to choose a real valued arrangement of the components lengths as a chromosome,each components length being a gene.According to biology principles of natural bee swarm,there are two populations,one population for global search,and another for local search.Only best one can crossover.The genetic operator includes tournament selection operator,order crossover operator,two-block-exchange mutation operator and restrain operator.The experiments results show that the bee swarm genetic algorithm approaches the theoretical optimal solution,its convergence rate is quick,and is efficient to solve one-dimensional cutting stock problem.
出处
《计算机技术与发展》
2010年第10期82-85,共4页
Computer Technology and Development
基金
黑龙江省2009年研究生创新科研资金项目(YJSCX2009-102HLJ)
关键词
一维下料问题
蜂群遗传算法
优化
最优交叉
抑制算子
one-dimensional cutting stock problem
bee swarm genetic algorithm
optimization
best one crossover
restrain operator
