基于线性回归的焦炭质量预测模型研究  

Research on a Linear Regression-based Coke Quality Prediction Model

作　　者：蔡成旺 潘芷莹 蔡孝庆 张小勇[1] 张代林[1] Cai Chengwang;Pan Zhiying;Cai Xiaoqing;Zhang Xiaoyong;Zhang Dailin(Anhui Provincial Key Laboratory of Clean Transformation and High Value Utilization of Coal,Anhui University of Technology,Anhui,243032)

机构地区：[1]安徽工业大学煤清洁转化与高值化利用安徽省重点实验室,安徽243032

出　　处：《当代化工研究》2023年第1期1-4,共4页Modern Chemical Research

摘　　要：对焦炭质量指标进行预测,是优化配煤结构,降低配煤成本,提高生产管理水平的重要举措。通过研究焦炭质量的影响因素,利用配合煤指标的灰分Am,d、挥发分Vdaf、粘结指数G、固定碳FCad、硫分Sm,t,d、胶质层指数(X、Y)和焦煤配比J作为自变量,分别进行一元或多元线性回归分析,得出的预测值同实测值相比较,预测精度符合要求,可用于指导实际生产。Forecasting coke quality is an important step to optimise the coal allocation structure, reduce coal allocation costs and improve production management. By studying the influencing factors of coke quality, using the blending coal index of Ash Am,d, Volatile Fraction Vdaf, Bond Index G, Fixed Carbon FCad, Sulphur Sm,t,d, Colloid Layer Index(X, Y) and Coking Coal Blending Ratio J as independent variables, respectively for single or multiple linear regression analysis, it is concluded that, in comparison with the predicted values with the measured values of prediction accuracy meets the requirements, can be used to guide the practical production.

关 键 词：配合煤指标 焦炭质量 预测模型 线性回归 

分 类 号：TQ[化学工程]