基于基因算法的焦炭质量预测模型的研究  

Study on predictive model for coke quality with algorithm based on gene

作　　者：周克城[1] 甘朝晖[1] 李高斌[1] 

机构地区：[1]武汉科技大学,武汉430081

出　　处：《燃料与化工》2011年第3期17-20,共4页Fuel & Chemical Processes

摘　　要：选取干燥无灰基挥发分(Vdaf)、胶质层最大厚度(Y)和炭化室高宽比(L/B)作为自变量,通过基于基因表达式的克隆选择算法对4种不同类型焦炉的生产数据进行分析,并建立焦炭质量的预测模型。对比采用该方法与回归分析得到的预测模型,该方法所得焦炭质量的预测模型的误差要远小于后者,能够更好地满足实际生产需求,为焦化厂快速准确地得到配煤方案提供了理论依据。By selecting dry ash-free volatile matter（Vdaf） ,maximum plastic zone thickness（Y） and height and width ratio of coking chamber（L/B） as independent variables,the production data of 4 different type coke ovens are analyzed through the clone selection algorithm based on gene expression,and the predictive model for coke quality is established.In comparison with the predictive models obtained with the method and regression analysis method,the error of the coke quality predictive model obtained by the method is far less than that obtained by regression analysis method,the model can meet the actual production requirement better and provides the theoretical base for getting coal blending ratio quickly and accurately for coking plant.

关 键 词：焦炭质量 GE-CSA算法 GEP编码 预测模型 

分 类 号：TQ520.1[化学工程—煤化学工程]