炼钢-连铸作业计划的遗传优化模型  

Optimization Model Based on Genetic Algorithms for Production Planning of Steelmaking and Continuous Casting Process

作　　者：朱道飞[1] 郑忠[1] 高小强[2] 

机构地区：[1]重庆大学材料科学与工程学院,重庆400045 [2]重庆大学经济与工商管理学院,重庆400044

出　　处：《钢铁》2008年第7期26-31,共6页Iron and Steel

基　　金：国家自然科学基金钢铁联合基金资助项目(50574110)

摘　　要：从提高炼钢-连铸作业计划可执行性的角度,研究了能适应动态生产环境变化的作业计划编制方法。在考虑连铸机的连浇约束、炼钢-连铸生产工艺要求和作业时间不确定性的条件下,建立了以最小化工位作业冲突时间和尽可能早的安排连铸机开浇时间为目标的炼钢-连铸生产作业计划优化模型,并构建了一种分段实数编码和基因分区交叉操作的改良遗传算法,将其与沿生产流程时间并行倒推算法相结合形成优化算法来求解模型。某钢厂8 h的炼钢-连铸作业计划编制案例表明:建立的模型与求解算法能较好地解决炼钢-连铸作业计划的时间不确定性优化问题,可快速生成炉次间作业无冲突的优化生产作业计划。In order to improve the performance of production planning of steelmaking and continuous casting the feasibility of adapting planning to dynamic environment was studied. An optimization model for production planning was established. In the model, the constraints of steelmaking continuous casting production,and sequence casting and the uncertainty of the operation time are taken into account to minimize the conflict time in operation and start casting as early as possible. An optimization algorithm was designed to solve the model, which combines an improved genetic algorithm with subset real number encoding, and divisional crossover operation and task allocation algorithm with parallel backward inferring along with production process. A case for eight hours production planning of steelmaking and continuous casting shows that the problems of time uncertainty in production planning can be solved and optimization production planning without conflict in operation can be made out quickly.

关 键 词：炼钢-连铸 作业计划优化 遗传算法 分段实数编码 

分 类 号：TF777[冶金工程—钢铁冶金]