加权马尔可夫链预测模型在降水量预测中的应用

Model of WMCP and its Application in the Prediction of Rainfall

基于雨季降水过程中存在大量的不确定性、非精确性的特点,阐述了马尔可夫链的理论及性质和加权马尔可夫链预测(WMCP)的思想,并应用有序聚类的方法(最优分割法)建立科学合理的状态分类标准。最后以川中丘陵区简阳市1974-2004年的降水量为例,运用1974年到2003年这30年降雨量所处状态来预测2004年降雨量所处的状态。整个分析、计算过程中的数据结果均是通过MATLAB程序来实现的,加权马尔可夫链预测方法计算方便,结果准确,为降雨量的预测分析提供了一个新的途径。

There are much uncertainty and imprecise characteristics in the course of rain in rainy season. This paper presented the theory and property of Markov chain and the idea of predicting based on WMCP; then applied sequential cluster method to set up a reasonable and scientific classification standard. In the end the data of the rainfall of Jianyang City in hill section of middle plain was choosed as an example, and predicted the rainfall of 2004 by analyzing rainfalls of 30 years from 1974 to 2003. All the data results of analysis and calculations in the paper are acquired by MATLAB program. WMCP is not complex, and the predicted rainfall matched the actual one. So the predicting method in this paper provided anatysts another way to predict rainfall.