炼钢厂生产调度排程的建模方法研究进展评述  

Review of research advances in modeling methods for production scheduling in steel mills

作　　者：郑忠[1] 张诗雨 王永周 张开 张开天 郭亮 高小强[3] ZHENG Zhong;ZHANG Shiyu;WANG Yongzhou;ZHANG Kai;ZHANG Kaitian;GUO Liang;GAO Xiaoqiang(School of Materials Science and Engineering,Chongqing University,Chongqing 400044,China;Shougang Jingtang United Iron and Steel Co.,Ltd.,Tangshan 063200,Hebei,China;School of Economics and Business Administration,Chongqing University,Chongqing 400044,China)

机构地区：[1]重庆大学材料科学与工程学院,重庆400044 [2]首钢京唐钢铁联合有限责任公司,河北唐山063200 [3]重庆大学经济与工商管理学院,重庆400044

出　　处：《钢铁》2023年第11期69-83,共15页Iron and Steel

基　　金：国家重点研发计划资助项目(2020YFB1712803);国家自然科学基金资助项目(51734004)。

摘　　要：炼钢厂生产调度是钢铁制造执行层车间级生产运行管控的核心技术。在以市场合同驱动的生产模式下,钢厂生产调度是多品种小规模订单与大规模制造执行之间矛盾协调的关键难点之一。针对炼钢厂生产调度问题,对涉及的制造流程、调度问题基本特征、数学建模方法、优化目标和约束条件等进行了归纳分析,重点围绕炼钢厂生产调度的建模技术展开讨论。在介绍钢厂调度主流的数学规划和系统仿真建模方法特点和研究进展的基础上,对相关商业软件在钢厂调度排程领域的应用情况,以及企业相关技术现状进行了讨论。尽管近些年,相关研究受到广泛的关注,取得了明显进展,但企业在线稳定应用尚难以见到,主要在于钢厂调度问题本身的复杂性,以及对问题认识还不够深入,导致建模假设的理想化,使得模型在处理大规模复杂钢厂调度问题时的方法过于简化、不能适应现实多场景的变化,存在应用局限。物联网、5G、工业互联网等新一代信息技术和人工智能技术等为制造业智能化发展提供了新技术和手段,通过对新技术在制造业调度领域的应用可能性的了解与分析,对钢厂生产调度领域技术发展进行了展望。感知、认知、决策和执行的闭环过程是智能制造的本质,为此,对信息技术发展背景下的钢厂生产调度领域重点需关注的问题进行了思考和展望。未来将借助IoT技术实现调度信息的实时感知、5G技术实现信息快速交互,工业互联网支撑钢厂生产调度系统的智能决策与优化控制,人工智能技术协助进行数据驱动的调度策略决策和关键参数优化,探索人工调度经验知识的规范表达与科学应用,以及基于人机协同的复杂钢厂的生产调度。该构想将为突破现有方法问题描述难、模型系统适应性差的难点和痛点问题提供新的解决方案。Steel mill production scheduling(SMPS)is a core technology for production operation control at the executive level of steel production.Under a contract-driven production mode,the SMPS method is one of the key points in reconciling the contradiction between diverse and small-scale orders and large-scale production.This paper provides an overview of the steel mill production process,basic characteristics of the SMPS problem,abstract description methods,optimization objectives and common constraints for this problem.Then,the modelling methods for SMPS are analyzed and discussed.Firstly,both the mathematical scheduling modeling method and the system simulation modeling method are discussed in detail by analyzing characteristics and current progress of research and application.The complexity of the SMPS and the insufficient understanding of the problem lead to the idealization of the modeling hypothesis,which result in a poor model adaptability.When dealing with large-scale complex SMPS problems,the over-simplified model fail to adapt to the changes of multiple scenarios in reality.New-generation information technologies,including IoT,5G,industrial internet and artificial intelligence,etc.provide new opportunities and challenges for the intelligent development of the manufacturing industry.By analyzing the potential application of new technologies in the production scheduling of manufacturing industry fields,an outlook on the technological development in the field of SMPS is made.The closed-loop process of sensing,cognition,decision making and execution is the essence of intelligent manufacturing.Therefore,considering the development of information technology,we think about its focus question and made a prognosis.In the future,real-time information awareness and the rapid interaction of information could be realized with the help of IoT and 5G technologies,respectively.And the industrial internet would support the intelligent decision-making and optimization control of the SMPS system.Additionally,artificial intellige

关 键 词：炼钢厂 生产调度 建模方法 工业互联网 人工智能 

分 类 号：TF758[冶金工程—钢铁冶金]