考虑时空约束的炼钢区域多行车协同调度优化研究与应用  

Research and application of collaborative scheduling optimization for multi-crane in steelmaking workshop considering spatial and temporal constraint

作　　者：夏建超[1] 王志远[2] 黄成永[1] 韩中洋 XIA Jianchao;WANG Zhiyuan;HUANG Chengyong;HAN Zhongyang(Steelmaking Plant,Shanghai Meishan Iron and Steel Co.,Ltd.,Nanjing 210039,China;Key Laboratory of Intelligent Control and Optimization for Industrial Equipment of Ministry of Education,Dalian University of Technology,Dalian 116024,China)

机构地区：[1]上海梅山钢铁股份有限公司炼钢厂,江苏南京210039 [2]大连理工大学工业装备智能控制与优化教育部重点实验室,辽宁大连116024

出　　处：《冶金自动化》2022年第2期80-89,共10页Metallurgical Industry Automation

摘　　要：多行车协同调度问题含有时空约束,是一个典型的多机多任务的复杂非确定性多项式难题,传统求解方法难以良好应对。鉴于此,建立了一种考虑时空约束的多行车协同调度问题模型。该模型以行车最少完工时长为优化目标,同时在时间和空间上分别考虑了行车的避碰约束和运行约束。提出了一种兼顾“群体智慧”和“个体认知”的改进遗传算法,对调度模型进行有效求解,基于真实数据的仿真实验结果表明,相较于现场人工调度,调度优化性能提高了65%。同时,将上述模型方法开发为应用软件,应用于国内某大型钢铁企业炼钢区域,进一步证明了方法的有效性和实用性。Multi-crane collaborative scheduling problem,which exhibits characteristics of multi-machine,multi-constraint and NP-hard with spatial and temporal constraint,is difficult to be solved by traditional methods.As such,a model takes the minimum completion time of the crane as the optimization goal and the collision avoidance along with operation restrictions in time and space as constraints is proposed.In order to efficiently and effectively solve this model,an improved genetic algorithm was proposed which integrates both swarm intelligence and individual cognition.It is demonstrated from the simulation results based on real-world data that the scheduling optimization performance is improved by 65%compared with manual scheduling.Furthermore,an application software developed based on the proposed model and method is successfully applied on a steelmaking workshop of a steel plant in China,which also manifested the effectiveness and practicability of this study.

关 键 词：炼钢区域 多行车调度 仿真模型 遗传算法 时空约束 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TF703[自动化与计算机技术—控制科学与工程]