基于递进式蚁群算法求解多目标汽车制造排程问题  

A Step-ACO Algorithm for the Multi-objective Car Sequencing Problem

作　　者：叶明[1] 王宁生[1] 

机构地区：[1]南京航空航天大学,南京210016

出　　处：《中国机械工程》2006年第14期1472-1476,共5页China Mechanical Engineering

基　　金：国家863高技术研究发展计划资助项目(863-511-910-403)

摘　　要：描述了多目标汽车排程问题的模型,提出了一个整体解决策略,即利用不同生产阶段间的缓冲区,通过基于改进的蚁群优化算法,实现多目标汽车队列优化以及有限柔性下的队列二次优化,递进式地求解该问题。提出运用蚁群算法解决以降低喷漆清洗成本和“重要选装件”使用率均衡为目标的汽车队列优化问题;设计了候选集—蚁群算法求解环形油漆车身缓冲区结构约束下的汽车队列二次优化问题。算例分析结果表明,提出的整体解决策略及算法具有有效性和优越性。This paper presented a model and the overall solutions for the multi--objective car sequencing problem. By the combination of multi--objective and single--objective sequencing optimization, we used the buffers between different production stages to solve this problem. First, we used ACO get the car sequence which satisfied the objectives of minimizing the costs of color purge and stability of high priority option usage rates. Then we designed Candidate-ACO to solve the CSP with limited flexibility under the round structure painted body storage. At last, data show that the proposed strategy and the algorithms are effective and useful.

关 键 词：汽车排程问题(CSP) 多目标优化 蚁群优化(ACO) 油漆车身缓冲区 

分 类 号：TP278[自动化与计算机技术—检测技术与自动化装置]