基于遗传算法和仿真分析的混合装配线平衡问题研究

Combining genetic algorithm and simulation analysis for mixed-model assembly line balancing problem

为兼顾混合装配线平均负荷平衡和瞬时负荷平衡,提出了一种综合运用遗传算法和仿真分析的混合装配线平衡问题的求解方法。首先,基于综合作业顺序图和多品种产品在每个作业任务上的平均作业时间,采用遗传算法求解混合装配线平衡问题,其优化目标是均衡各工作站平均作业时间;然后,对遗传算法求解的一组较优解,从瞬时负荷平衡方面进行仿真分析,其优化目标是最大化各工作站利用率;最后,综合两个优化目标确定混合装配线平衡问题的最优解。通过算例分析,验证了求解方法的有效性。

To seek the optimal solution to balance average workload and instantaneous workload in mixed-model assembly line balancing problem, an approach of combining genetic algorithm and simulation analysis was proposed. Firstly, a genetic algorithm was applied to solve the mixed-model assembly line balancing problem based on the combined precedence diagram and average processing time for each task of the models; the objective was to minimize the difference of average workload among the workstations. Then, the simulation analysis was carried out based on the optimal or sub-optimal solutions provided by the genetic algorithm to analyze instantaneous workload balance; the objective was to maximize the weighted workstation efficiency. Finally, the optimal solution to the mixed-model assembly line balancing problem was defined by two optimization objectives. Effectiveness of the approach was proved by case study.