面向节能的时间窗约束下炼钢生产调度与物流优化

With the requirement of energy saving and emission reduction become more and more stringently, how to lessening temperature dropping of ladles to save energy is an urgent problem in steelmaking production which needs to be sloved. Lessening waiting time of ladles by scheduling and logistics optimization is an effiective way to slove this problem. Therefore, this project will make the fundamental scientific research of steelmaking scheduling and logistics optimizatin in view of time window constraints. 4 parts are inclued in this project. The first part is studying steelmaking scheduling problem. Steelmaking scheduling considering time window constraints is classified as flow shop scheduling problem with waiting time window constraints(FSPWTW). The mechanism of FSPWTW problem will be probed and the schedule law will be analyzed. A new schedule method will be presented which combined group adjustment with neighborhood search. The second part is studying crane scheduling problem by simulation. The crane schedule law will be studied by simulation and a crane simulation scheduling framework will be presented. Concurrent simulation method will be used to speed-up simulaiton and scheduling. The third part is integration of scheduling and logistics. An integrated model will be built and a reversed cycle optimization method will be presented to optimizing the upper bound of time window constraints. The fourth part are experiments and verification. The experiment system of scheduling and logistics optimization will be built and some steelmaking cases will be used to test and verify all the work above. This project may discover the scheduling law of waiting time window and build an effiective way to integrate scheduling and logistic optimization. This project may also help steel enterprises to save energy. Therefore, this project has great scientific meaning and engineered value.

在节能减排要求日益严苛的背景下，减少钢包温降以追求节能是炼钢生产需要迫切解决的问题。通过生产调度和物流优化以减少钢包等待时间是解决这一问题的有效途径。为此，本项目从物流时间窗约束角度对炼钢生产调度及物流优化进行基础科学研究，包括：将考虑物流时间窗的炼钢生产调度问题归结为等待时间存在时间窗约束的流水调度问题，探索该调度问题的约束机理，分析时间窗约束下的调度规律，提出一种约束联动结合邻域搜索的调度方法；在仿真环境下研究天车调度规律，建立基于Agent的天车仿真调度架构，提出使用并行仿真方法实现快速仿真调度；建立生产调度与物流的集成模型，提出使用反向循环优化的方法对时间窗约束上限进行优化；最后建立生产调度与物流优化实验系统，并结合炼钢生产实例对上述工作进行验证。本项目在揭示物流时间窗约束下的调度规律，探索生产调度与物流的集成优化方法，提高钢铁企业节能减排水平等方面具有理论意义和工程应用价值。

在节能减排的背景下，针对炼钢生产调度与物流优化进行研究。通过生产调度和物流优化减少钢包热等待时间，保障生产效率的同时避免不必要的能耗。主要完成以下工作：1）以钢包的物流时间为下限，以钢包最大等待时间为上限，形成考虑等待时间窗的流水调度（Flowshop）问题。将此问题进一步细分为置换流水问题（PFSP，Permutation flowshop）和非置换流水问题（NPFSP,Non-Permutation flowshop），分析了在不同机器数量情形下和不同时间窗作用下的PFSP和NPFSP支配规律。进一步针对复杂的多机调度情形，分别提出用改进的迭代贪婪算法（IG, Iterated Greedy）和非置换邻域调度启发式算法（NNS,Neighborhood non-PFSP searching heuristic）予以求解。实验表明，分别获得相近质量FPSP和NPFSP调度解，在计算时间上IG方法约为遗传算法的11%，NNS方法约为memetic算法的1%。2）在时间窗约束下，考虑天车冲突，形成天车吊运的物流优化问题。针对该问题建立了约束调度数学模型并结合工业要求将天车冲突分类。进一步提出基于RESTART（Repetitive Simulation Trials After Reaching Thresholds）的并行仿真调度方法。实验表明，本文方法相对范德兰德工业公司使用的调度方法，调度解可优化30%。3）在时间窗约束下，形成生产与物流的集成优化问题。针对该问题提出集成优化框架并建立了原型实验系统。实验结果表明，该系统可以发现紧时间窗约束下的可行解。以上方法专门针对等待时间窗的流水调度和天车调度问题而设计，从而获得了较好的计算效率，适合炼钢工业应用。此外，在研究过程中发现，上述方法可拓展到自寻址小车存取系统。上述研究遵循建模—方法开发—实验分析的主线，分别实现了炼钢生产调度、物流调度以及集成优化。对于丰富优化理论、推动优化软件的工业应用具有重要意义。