马尔可夫链预测方法的统计试验研究

Study on statistical experiments of markov chain prediction methods

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的预测理论中得到了广泛的应用。各种马尔可夫链预测方法在有关应用实例中似乎效果都不错,然而,实际上他们的预测精度是有差别的。本文通过统计试验手段对这三种马尔可夫链预测方法进行了比较研究,结论表明:WMCP方法预测精度最高,SPMCP方法预测精度次之,ADMCP方法预测精度相对来讲最差。

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) were widely used in various Hydrology series.The results of various Markov chain prediction methods in real instances looked all sound.But at matter of fact,the prediction accuracy of three kinds of Markov chain prediction methods are different.Which are studied based on statistical experiments in the ...