滑动平均-权马尔可夫链模型在降水预测中的应用

基于降水过程存在大量不确定性的特点，以徐州市53年的降水资料为实例，通过均值标准差分级法建立降水序列的分级标准，采用规范化的各阶自相关系数为权重，用滑动平均的权马尔可夫链模型，通过状态转移概率矩阵预测未来时段的降水状态，并应用级别特征值计算具体的降水量，结果表明该方法预测精度较高，结果与实际情况相吻合．

Applieation of Sliding Mean-Weighted Markov Chain Model on Rainfal Prediction

Based on the uncertain characteristics in precipitation procession, the classification standard of precipitation serial was set through the mean and standard deviation of information series, the normalized different autocorrelation coefficient was used as weight, the moving average-weighted-Markov chain was used to predict precipitation state in a future period with state transition probability matrix, then the particular precipitation was calculated according to the class designated state, while the precipitation material in 53 years of XuZhou country was used as a model. It is concluded that the result of prediction by the present method agrees with the reality, and the precipitation accuracy was satisfied.