基于随机森林的中长期降水量预测模型研究

为解决降水预测中存在的非线性问题,避免传统人工智能方法的过拟合弊端,提高中长期降水预测精度,引入了随机森林算法,通过优选预报因子,分别构建了年、月降水预测模型,并应用南京市1951~2013年降水系列及水文气象因子系列,验证所建模型的适用性。结果表明,随机森林模型预测精度较高、稳定性好、泛化能力强,能有效预测年、月降水量;与BP神经网络模型和支持向量机模型相比,随机森林模型效率更高、性能更优,尤其适用于大样本的逐月降水量预测中。

Research of Medium and Long Term Precipitation Forecasting Model Based on Random Forest

In order to solve the nonlinear problems of precipitation forecast, the random forest (RF) algorithm was introduced to avoid the over fitting problem of traditional artificial intelligence and improve the accuracy of medium and long term precipitation forecast. Based on the random forest, a prediction model was constructed by using the optimal selected predictors, and the applicability of the model was verified by the precipitation series and hydrological and meteorological factor series of Nanjing from 1951 to 2013. The results show that the random forest model has high precision and good generalization, which makes RF an efficient precipitation prediction method. Compared with the BP artificial neural networks and support vector machine model, the random forest model is better than the other two models, especially in large samples monthly precipitation forecast.