基于自组织云演化算法的铁水运输调度理论与方法研究

In iron and steel production system, molten iron transportation is the process between iron and steel making procedures, which has the important role. With the enterprise re-organization, the molten iron transportation scheduling in the distributed manufacturing environment becomes more and more complex, which needs high performance optimization algorithms. In this project, we investigate the problem characteristics of molten iron transportation. Then, the hybrid flow shop with flexible operation selection in distributed manufacturing (HFS-FD) is firstly proposed. We combine the problem features, i.e., distributed, multi-constraints, multi-objective characteristics, to design the skeletons, and the fitness landscape for HFS-FD problems. Then, we consider both the big data analysis under the off-line state and rescheduling approaches under the on-line state, to investigate the rescheduling methods. Next, we design componential, parallel, self-organization discrete evolutionary algorithm framework considering the problem features. Last, we verify and develop the proposed evolutionary optimization algorithm on the realistic production data. The project aims to provide a batch of theory with project application values in many aspects, such as operation route selection approaches and rescheduling methods in HFS-FD, distribution and multi-objective mechanism, and cloud evolutionary algorithm. The research achievements will serve the iron and steel enterprise and improve the management level and competition capabilities.

钢铁生产中铁水运输过程处于炼铁与炼钢工序之间，起着承上启下的作用。随着钢铁企业重组，分布式制造环境下铁水运输调度问题将更加复杂，亟待高性能优化算法求解。本项目针对铁水运输过程，提炼出分布式柔性工序选择的混合流水车间（hybrid flow shop with flexible operation selection in distributed manufacturing, HFS-FD）调度问题。研究分布式、多约束、多目标HFS-FD问题骨架和适应度地貌，设计离线大数据预测分析和在线重调度相结合的策略，构建问题相关的组件化、并行化、自组织云演化算法，基于实际生产数据验证和发展上述演化算法。项目旨在HFS-FD工序路线选择、重调度策略、分布式及多目标处理、云演化算法设计等方面给出一批具有工程应用价值的理论成果。研究成果可望直接服务于我国钢铁生产企业，促进其生产管理水平和市场竞争力的提高。

钢铁生产中铁水运输处于炼铁与炼钢工序之间，起着承上启下的作用。随着企业重组，分布式制造环境下铁水运输调度问题将更加复杂，亟待高性能优化算法求解。本项目针对铁水运输过程，提炼出HFS-FD(Hybrid Flowshop with flexible operation selection in distributed manufacturing, 分布式柔性工序选择的混合流水车间)调度问题。研究分布式、多约束、多目标的问题骨架和适应度地貌，设计离线大数据分析和在线重调度相结合的策略，构建问题相关的组件化、并行化、自组织云演化算法，基于实际生产数据验证和发展上述演化算法。项目在HFS-FD工序路线选择、重调度策略、分布式及多目标处理、云演化算法设计等方面给出了一批具有工程应用价值的理论成果，共完成论文35篇，其中SCI论文31篇，EI论文4篇，授权发明专利7项，申请发明专利12项，培养硕士研究生12名，举办国内会议1次，参加国际会议并作分组报告13次，获省部级奖励1项，省高校科研奖励1项。研究成果可望服务于我国钢铁生产企业，促进其生产管理水平和市场竞争力的提高。