摘要
针对生产装配线平衡问题,提出一种改进的遗传算法。算法采用缩放适应度法、随机普遍取样的选择策略、线性可变的杂交和变异算子。使用PB语言实现了这一应用平台,给出了系统的功能结构图和主要的数据结构,并结合实例给出了ALB-2问题的解决方案。实例对比证明,改进的算法很好地解决了简单遗传算法易早熟的问题,大大改善了简单算法的性能。
An improved genetic algorithm is present for ALB problem, which uses a scaling fitness method, a stochastic universal sampling selection way, linear-changed crossover and mutation probabilities. The platform is achieved by PowerBuilder, the architecture and main data structure are also brought forth. The solution of ALB-2 problem is described by an example. The pre-maturity problem of simple genetic algorithms is resolved and the performance of simple genetic algorithms is increased greatly.
出处
《计算机工程与设计》
CSCD
北大核心
2006年第20期3731-3733,3736,共4页
Computer Engineering and Design
基金
国家863高技术研究发展计划基金项目(2003AA414011)
湖北省教育厅科学研究计划基金项目(B200511008)
关键词
装配线平衡问题
遗传算法
任务偏序图
缩放适应度
随机普遍取样
assembly line balancing
genetic algorithm
ordered graph of tasks
scaling fitness
stochastic universal sampling
