基于改进差分进化算法的多目标第Ⅰ类装配线平衡问题研究  

Improved Differential Evolution Algorithm for Multi-objective Assembly Line Balancing Problems of Type Ⅰ

作　　者：闫琼[1] 张海军[1] 张国辉[1] 刘元朋[1] 

机构地区：[1]郑州航空工业管理学院,郑州450015

出　　处：《科技通报》2017年第11期197-200,共4页Bulletin of Science and Technology

基　　金：国家自然科学基金青年基金(51705472);航空科学基金项目(2015ZG55018);河南省高等学校重点科研项目(18A460033);校青年科研基金项目(2016053001)

摘　　要：为求解多目标第Ⅰ类装配线平衡问题(MOABLP-Ⅰ),提出了一种改进的差分进化算法(IDEA)。该算法优化目标包括最优工位数,线生产效率和工位载荷波动。采用基于优先权的编码方法使得个体解码后总满足装配线约束关系,设计了自适应双变异策略和新型交叉操作算子使算法适应离散优化问题,引入"精英保留"机制增强算法逃离局部最优的能力。通过测试问题集的验证,并比较了基本差分进化算法和离散型差分进化算法,结果表明IDEA在求解大规模MOABLP-Ⅰ上质量最优。To solve the multi-objective assembly line balancing problem of type 1, an improved differential evolution algorithm(IDEA) was proposed. The objectives considered here ware minimization of number of workstations, maximization of assembly line efficiency, and minimization of workload variation. IDEA employed the method of indirect coding based on priority, which is to ensure the decoding particles satisfy the precedence constrains of assembly operations. The self-adaptive double mutation strategy and new crossover operator were developed in IDEA. In order to avoid being stuck in the local convergence, the'elite-preserving mechanism'was introduced in the select operator. The experimental results showed that the IDEA was superior to the basic DEA and the discrete DEA.

关 键 词：第Ⅰ类装配线平衡问题 多目标优化 差分进化算法 离散 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]