基于 MGASA 的装配车间物流协同优化方法研究

针对乘用车发动机装配车间内处理大规模订单排产和产品配送调度方案存在求解时间长、效率低、协同优化效果不明显的问题,文章提出一种基于改进遗传模拟退火算法(modified genetic algorithm and simulated annealing, MGASA)的装配车间物流协同优化方法。分析多品种小批量面向订单式生产的乘用车装配车间物流的特点,确定优化目标为最小化客户期望时间、提前延迟成本和物流配送成本;针对问题特征提出装配订单生产配送调度的优先级判定规则和4类特征指标以便进行问题编码和适应度计算,且在同一温度下多次进行种群迭代进化和淬火操作,扩大可行解的邻域范围,以期获得全局最优解,得到装配车间内的生产配送调度方案;最后在不同规模的数据集上进行实例验证。实验结果表明,该方法可达到较高的求解效率,实现乘用车装配车间物流协同优化调度方案的快速制定,具有一定的应用价值。

To address the challenges of prolonged solution times, low efficiency, and inadequate collaborative optimization in handling large-scale order scheduling and product distribution scheduling in passenger car engine assembly workshops characterized by multi-variety, small-batch, and order-oriented production, this paper proposes a collaborative optimization method based on modified genetic algorithm and simulated annealing (MGASA). The optimization model aims at minimizing customer expectation time, earliness-tardiness costs, and logistics distribution costs. A priority rule and four categories of characteristic indicators are adopted for problem encoding and fitness evaluation. Iterative population evolution and neighborhood solution generation are repeatedly executed at the same annealing temperature to expand the neighborhood of feasible solutions, facilitating the attainment of a globally optimal solution. Consequently, the production and logistics scheduling schemes for the assembly workshops are determined. Experimental validation conducted on datasets of various sizes demonstrates that the proposed method significantly enhances solution efficiency and supports rapid decision-making in collaborative logistics scheduling, highlighting its practical applicability to passenger car engine assembly workshops.