摘要
拆卸线平衡问题的优化涉及多个目标。为克服传统方法在求解多目标拆卸线平衡问题时不能很好地处理各子目标间冲突及易于早熟等不足,提出了一种多目标细菌觅食优化算法。该算法采用Pareto非劣排序技术对种群进行分级,并结合拥挤距离机制评价同级个体的优劣。为提高算法收敛性能,在趋向性操作结束后引入精英保留策略保留优秀个体,并采用全局信息共享策略引导菌群不断向均匀分布的Pareto最优前沿趋近。通过不同规模算例的对比分析,验证了算法的有效性与优越性。
The optimization procedure of DLBP involves dealing with multiple objectives. Traditional algorithms could not handle the conflict between objectives properly and might get local optimum prematurely. To hedge against these shortcomings, this paper proposed a Pareto based multi-objective bacteria foraging optimization algorithm. The algorithm used a Pareto non- dominated sorting operator to grade the bacterial population. For those solutions which belong to the same grade, it adopted a crowding distance operator for the second rank. After chemotaxis phase, the algorithm introduced an elitism preservation stra- tegy so that it would improve the convergence performance of the proposed algorithm. Furthermore, the algorithm used a global information sharing strategy to guide the bacterial population searching toward the well distributed Pareto optimal front. Compu- tational comparisons of different size DLBP instances demonstrate the validity and superiority of the proposed algorithm.
出处
《计算机应用研究》
CSCD
北大核心
2016年第11期3265-3269,共5页
Application Research of Computers
基金
国家自然科学基金资助项目(51205328
51405403)
国家教育部人文社会科学研究青年基金资助项目(12YJCZH296)
四川省应用基础研究计划项目(2014JY0232)
关键词
多目标优化
PARETO解
拆卸线平衡
细菌觅食优化算法
multi-objective optimization
Pareto solutions
disassembly line balancing
bacteria foraging optimization algorithm
