求解分布式混合流水线调度问题的改进双层嵌套式遗传算法

针对分布式混合流水线生产的生产调度问题,模拟实际排产中的排产到线和排产到时的排产策略,提出了基于改进双层嵌套式遗传算法的两层优化模型。外层依据流水线分配平衡和准时交货等基本原则总体上解决生产订单在流水线之间的分配问题,内层以最小生产时间为主要目的求解流水线的生产订单生产次序问题。考虑到双层嵌套式遗传算法的时间复杂性,基于模糊逻辑理论设计了一种模糊控制器来动态调整遗传算子,并采用主动检测停止方法,提高算法效率。使用某空调工厂的实际生产数据验证了算法的可行性、计算结果的准确性及排产策略的有效性,为高级计划与排程(Advanced Planning and Scheduling,APS)中大规模复杂供应链调度问题提供了可借鉴的方法。

A scheduling algorithm for distributed hybrid flow-shop production scheduling problem

For distributed hybrid flow-shop production problem,proposed a two-layer optimization model based on improved double nested genetic algorithm.This model aims at simulating the strategy of the planning and scheduling of actual production.According to the principle of assembly line distribution balance and on-time delivery,the outer layer generally copes with the assignment problem of manufacturing tasks among different assembly lines.Furthermore,the inner layer deals with the production sequence problem of manufacturing tasks with the minimizing production time purpose.Considering the time complexity of the two-layer genetic algorithm,a fuzzy controller is designed based on the fuzzy logic theory to dynamically adjust the genetic factor,and an active detection method is used to improve the efficiency of the algorithm.By applying the production data of an air conditioning plant with our algorithm,this proposition has been verified as an effective and practicable approach for large-scale complex supply chain scheduling problems in APS.