考虑生产顺序依赖的离散批量调度问题研究
Research on Discrete Lot-sizing and Scheduling Problems with Production Sequence-dependence
张鋆 1唐昕 1郑澜波1
作者信息
- 1. 武汉理工大学交通与物流工程学院,武汉 430063
- 折叠
摘要
离散批量调度问题(DLSP)是一类特殊的生产规划问题.在该问题中,生产时间被分割成离散的片段,每个时段以全产能生产一种产品,除库存和生产成本外,还存在依赖于产品生产顺序的切换成本.解决此类问题需要同时决定批量的规模和批次顺序,这对有效地规划生产批次和调度,提高生产效率,控制成本至关重要.作者以约束规划为手段,研究单产线多产品的DLSP问题,在Fleischman[3]将问题转换为带时间窗的旅行商问题(TSPTW)的基础上,建立约束规划模型;针对模型在约束变量上联系并不紧密的情况,运用基于边际成本的域削减技术,极大缩减变量值域,随后进行不同规模的数值实验,验证模型和算法的有效性.实验结果表明:算法能够在60 s内有效提高初始解质量.
Abstract
The Discrete Lot-sizing and Scheduling Problem(DLSP)is a special class of production planning problem.In this problem,the production time is divided into discrete segments,and each time slot produces a product at full capacity,with changeover costs that depend on the production sequence of the products,in addition to inventory and production costs.Solving such problems requires deciding both the lot size and the lot sequence,which is crucial for efficiently plan-ning production lot sizing and scheduling,improving production efficiency,and controlling costs.The authors study the DLSP problem with multiple products on a single production line by means of constraint programming,and establish a constraint programming model based on Fleischmann[1]converting the problem into the traveling salesman problem with time windows(TSPTW);for the case that the model is not closely linked in the constraint variables,the marginal cost-based domain filtering technique is applied to greatly reduce the value domains of the variables.The domain filtering tech-nique(cost-based domain filtering)is applied to greatly reduce the domain of variable values,and then numerical experi-ments of different scales are carried out to verify the effectiveness of the model and the algorithm.The experimental re-sults show that the algorithm can effectively improve the initial solution quality within 60 seconds.
关键词
离散批量调度/约束规划/TSPTW/约束传播Key words
discrete lot-sizing and scheduling/constraint programming/TSPTW/constraint propagation引用本文复制引用
出版年
2024