基于齐次马尔可夫链的径流预测模型——以黄河花园口监测站为例  

Runoff predictionbased on homogeneous Markov chains——Taking the Huayuankou station on the Yellow River as an example

作　　者：焦建利[1] 李伟 李梦涛 刘畅 Jiao JianLi;LI Wei;Li MengTao;Liu Chang(Shanghai University of Medicine&Health Science,Shanghai 201318;School of Mathematics&Physics,Hebei University of Engineering,Hebei Handan 056038,China)

机构地区：[1]上海健康医学院,上海201318 [2]河北工程大学数理科学与工程学院,河北邯郸056038

出　　处：《南方农机》2020年第1期55-56,58,共3页

摘　　要：本文以周为时间尺度,以黄河花园口站为研究对象建立了马尔可夫链径流预测模型。模型以频数法构建转移概率矩阵,以极限分布作为初始分布。结果表明,预测效果较以年或月为时间尺度的传统模型精度更高。Based on the Markov chains by weeks,a runoff prediction model for Huayuankou station on the Yellow River is established.On the process of building this model,the transition matrix is established by frequency and the initial distribution is determined by limiting distribution.Finally,compared withthe obtained models,the result shows that the prediction accuracy of this Markov chains model is better than the prediction models with time scales of year or month.

关 键 词：马尔可夫链 径流预测 花园口 

分 类 号：O213.9[理学—概率论与数理统计]