Abstract: Based on considering the fluctuations of switching product,for solving mixed-model assembly line balancing problem,an improved Intelligent Water Drop (IWD) algorithm was proposed,which integrated three optimization objectives-workstation number,workload balance and task relatedness.The node metastasis rule of IWD algorithm was improved by adding largest probability leading rule and random search rule.The method of Pareto dominance was used to obtain frontier solution set and provide a heuristic value for each particle,and all particles were conducted global update to enhance global search ability according to the heuristic value.Through the experiment of standard test problems,the results showed that the improved IWD algorithm could solve the multi-objective mixed-model assembly line balancing problem more effectively than genetic algorithm.

Key words: mixed-model assembly line, intelligent water drop algorithm, Pareto dominant, task relatedness

摘要： 为简化混合装配平衡问题的求解,进而提高装配线的生产效率,在兼顾产品切换引起负荷波动的基础上,综合工作站数、工作负荷平衡和任务关联度三个优化目标,提出一种求解多目标混合品种装配线平衡问题的改进型IWD(intelligent water drop)算法。对IWD算法的节点转移规则进行改进,加入最大概率引导规则和随机搜索规则;采用Pareto占优的方式对解进行分层以获得前沿解集,并根据分层结果给每个粒子提供一个启发值,依据启发值实施全局更新,增加算法的全局搜索能力;通过测试各种标准问题,验证了改进型IWD算法比遗传算法的求解速度更快、效率更高。

关键词: 混合品种装配线, IWD算法, Pareto占优, 任务关联度