Canopy-Kmeans聚类和组合优化的铁矿预配料智能调度

发布时间：2018-03-01 19:23

本文关键词： 铁矿预配料有限下料槽 Canopy-Kmeans算法组合优化智能调度　出处：《控制理论与应用》2017年07期 　论文类型：期刊论文

【摘要】：铁矿预配料的原料种类繁多、化学成分差异较大,且下料槽个数有限、生产约束多,原料下料次序难以确定.针对该配料调度难题,本文提出了一种基于聚类算法和组合优化的铁矿混匀过程预配料智能调度方法.分别根据原料成分中SiO_2,TFe含量的差异,采用Canopy-Kmeans聚类方法进行两次聚类,然后综合考虑各项约束条件,利用融合专家规则的组合优化和小范围穷举思想对聚类结果进行组合与排序,得到原料共槽方案与共槽下料次序,以保证在有限下料槽的情况下配完所有原料,且配得的混匀料化学元素含量始终尽可能稳定.经我国某钢铁厂实际生产数据验证,所提方法与现有人工计算方法相比,大幅缩减了运算时间,且矿物化学元素指标的波动小,具有实用价值.
[Abstract]:There are many kinds of raw materials for iron ore pre-proportioning, big difference in chemical composition, limited number of feeding tanks, many production constraints, and difficult to determine the order of raw materials. This paper presents an intelligent scheduling method for iron ore blending process based on clustering algorithm and combination optimization. According to the difference of SiO2TFE content in raw material composition, Canopy-Kmeans clustering method is used to cluster twice. Then considering the constraint conditions synthetically, using the combination optimization of fusion expert rules and the idea of small range exhaustive, the clustering results are combined and sorted, and the scheme of common trough and the order of cutting stock in common trough are obtained. In order to ensure that all the raw materials are finished under the condition of limited blanking tank, and the chemical element content of the mixed material is as stable as possible, the method is verified by the actual production data of a steel and iron plant in our country, and compared with the existing manual calculation method. The calculation time is greatly reduced, and the fluctuation of mineral chemical element index is small, so it is of practical value.
【作者单位】：中南大学信息科学与工程学院;
【基金】：国家自然科学基金重大项目课题(61590921);国家自然科学基金资助项目(61273187) 中南大学中央高校基本科研业务费专项资金(2017zzts135)资助~~
【分类号】：TD921.6;TP301.6

【相似文献】