马氏链点值预测法和时序分析预测法的比较研究

Statistical Experiment Study on Markov Chain Point Value Prediction Method and Time Series Snalysis Prediction Method

近年来,基于绝对分布的马尔可夫链预测(ADMCP)方法、叠加马尔可夫链预测(SPMCP)方法和加权马尔可夫链预测(WMCP)方法在各种水文序列预测理论中得到了广泛的应用。然而,实际生活中仅仅预测出状态区间是不够的,文章给出一种基于马氏链状态预测方法的点值预测方法,并将其与普遍接受的时间序列分析预测方法进行了基于统计试验的比较分析,得出了该方法在水文序列预测中较优的结论。

In recent years, the Markov chain prediction method based on absolute distribution （ADMCP）, the Markov chain prediction method based on probability summation （SPMCP） and weighted Markov chain prediction method （WMCP） have been widely used in various hydrology series. However, in real life simply predict state interval is not enough, this paper presents a point value prediction method based on markov chain state prediction method, and made a comparative analysis with the generally accepted time series analysis forecasting method based on statistical experiment method, we obtain the conclusion that the point value prediction method is optimal in hydrological sequences.