基于马尔可夫理论的气固分选流化床密度的预测  

Density prediction of gas-solid fluidized bed for separation base on Markov theory

作　　者：骆振福[1] 陈尚龙[1] 赵跃民[1] 陈增强[1] 唐利刚[1] 

机构地区：[1]中国矿业大学化工学院,江苏徐州221116

出　　处：《煤炭学报》2011年第1期105-109,共5页Journal of China Coal Society

基　　金：国家自然科学基金委员会创新研究群体科学基金资助项目(50921002);国家自然科学基金资助项目(90510002);国家高技术研究发展计划(863)资助项目(2007AA05Z318);中国矿业大学中央高校基本科研业务费专项资金资助项目(2010ZDP01A06)

摘　　要：运用加权马尔可夫链模型对不同时刻气固分选流化床密度进行预测,结果表明:从开始记录流化床密度起,气固分选流化床在60 min时密度所处的状态最有可能位于[1.784 702,1.805 832)g/cm3,在62 min时最有可能位于[1.803 719,1.826 962)g/cm3;而实际测得气固分选流化床在60 min时密度为1.794 g/cm3,在62 min时密度为1.806 g/cm3。对加权马尔可夫链模型的预测精度进行分析,得到其预测偏差均值为0.262 879,预测偏差均方差为0.341 727。可见,利用加权马尔可夫链模型可以实现气固分选流化床密度的预测,预测精度较高。The density of gas-solid fluidized bed for separation at different time was forecasted by a weighted Markov chain model.The results show that from the time began to register,the density of gas-solid fluidized bed at 60 min will be[1.784 702,1.805 832） g/cm3,and that at 62 min will be[1.803 719,1.826 962） g/cm3.The measurement results of bed density are 1.794 g/cm3 at 60 min and 1.806 g/cm3 at 62 min.The prediction precision of weighted Markov chain model was analyzed.The results show that the mean of prediction deflection is 0.262 879 and the mean square deviation of prediction deflection is 0.341 727.It can be seen that the density of gas-solid fluidized bed for separation can be forecasted by a weighted Markov chain model with higher prediction precision.

关 键 词：气固分选流化床 马尔可夫链 预测模型 

分 类 号：TD922[矿业工程—选矿]