基于树回归的石灰竖窑煅烧质量预测模型  

Prediction model of calcination quality of lime shaft kiln based on tree regression

作　　者：王刚 刘前 程伟娇 何飞 藏庆涛 张友臣 曲长乐 Wang Gang;Liu Qian;Cheng Weijiao;He Fei;Zang Qingtao;Zhang Youchen;Qu Changle(Jianlong Xilin Iron and Steel Co.,Ltd.,Yichun 153025,China;National Sintering and Pelletizing Equipment System Engineering Research Center,Changsha 410205,China;Zhongye Changtian International Engineering Co.,Ltd.,Changsha 410205,China)

机构地区：[1]建龙西林钢铁有限公司,伊春153025 [2]国家烧结球团装备系统工程技术研究中心,长沙410205 [3]中冶长天国际工程有限责任公司,长沙410205

出　　处：《耐火与石灰》2024年第5期11-17,共7页Refractories & Lime

摘　　要：为了提高石灰竖窑生产过程的自动化控制精确性,建龙西林钢铁厂提出采用树回归动态预测成品CaO含量;针对控制参量多,耦合关系复杂的问题,采用相关性分析的方法优化参量,由此提高随机森林回归算法预测的可靠性;通过9个特征参量样本数据测试表明,树回归模型可以很好地适应石灰煅烧多因素、非线性的工业场景,预测结果精度满足工业控制要求。In order to improve the accuracy of automatic control of the production process of lime shaft kiln,Jianlong Xilin Iron and Steel Plant proposed to use tree regression to dynamically predict the CaO content of finished products.Aiming at the problem of many control parameters and complex coupling relationship,the correlation analysis method is used to optimize the parameters,so as to improve the reliability of the prediction of the random forest regression algorithm.Through the test of 9 characteristic parameter sample data,it is shown that the tree regression model can adapt well to the multi-factor and nonlinear industrial scene of gray stone calcination,and the accuracy of the prediction results meets the requirements of industrial control.

关 键 词：竖窑 石灰 过程控制 树回归 预测 

分 类 号：TQ177.26[化学工程—硅酸盐工业]