马尔可夫链在水库入库径流状态预测中的应用  

The Application of Markov Chain to Forecasting Reservoir Inflow State

作　　者：王永兵[1] 胡小梅[1] 彭丹芬[1] 陈崇德[1] 

机构地区：[1]湖北省漳河工程管理局,湖北荆门448156

出　　处：《水电与新能源》2011年第4期18-21,共4页Hydropower and New Energy

摘　　要：依据漳河水库1963-2008年入库年径流量资料,应用均值标准差法建立5级分级标准。针对径流量为相依随机变量的特点,以各阶自相关系数为权重,运用马尔可夫链模型预测未来一年的入库径流量状态。结果表明：该方法直观、预测准确、计算简便,为区域径流量的中长期预测提供了新的分析途径。Based on the annual run off data form 1963 to 2008 in Zhanghe River Reservoir,five grading standards are established by use of deviation from mean.Aiming at the runoff characteristics of dependent random variables,and taking the order of autocorrelation coefficients as weights,Markov chain model is used to predict the following year runoff state.The results show that the method is intuitive,accurate in prediction and simple in computation,which provides a new way to predict mid-and long-term runoff.

关 键 词：自相关系数 马尔可夫链 径流量 预测 漳河水库 

分 类 号：TV121[水利工程—水文学及水资源] P333[天文地球—水文科学]