基于改进马尔可夫链的径流预测模型

Runoff Predicting Model Based on Improved Markov Chain

径流预测在区域水资源规划中发挥着重要作用。基于径流过程不确定性、不精确性的特点,利用资料系列的均方差方法把径流序列分为不同的状态。在此基础上,采用统计方法,建立转移概率矩阵,以规范化的各阶自相关系数为权重,用加权的马尔可夫模型预测径流状态,并根据模糊集理论中的级别特征值预测径流量。最后把改进后的马尔可夫预测模型应用于宝鸡市的渭河径流丰枯状态及年径流量的预测中。1998￣2000年径流量预测的相对误差分别为3.64%,7.50%和1.00%,结果较好。

Runoff prediction plays an important role in water resources planning. Based on the uncertainty and inaccuracy characteristics, runoff series could be divided into different states via mean square deviation method of data series. Transition probability matrix was obtained by using statistical method. Standardized self-correlative coefficients based on the special characteristics of correlation among the historical stochastic variables were regarded as weights. The method of Markov chain with weights was used to predict the future runoff state, and the level characteristics value of fuzzy sets was used to predict the concrete value of runoff. Furthermore, the improved model was applied to predict annual runoff of the Wei River in Baoji as an example. The relative error of annual runoff prediction from 1998 to 2000 were 3.64% ,7.50% and 1.00%respectively , the results were satisfactory.