Abstract：Based on deeply analyzing the meteorological data of the local area in the south,a mixed multi-neural-network model （Probabilistic Neural Network and Radial Basis Function Neural Network model,referred to as PNN and RBF model） is proposed in this paper,which is used to identify the type of rainfall and forecast the rainfall quantity.The model is composed with one PNN neural network and two RBF neural networks,in which,the former （the PNN neural network） is used to classify the type of rain,and two different RBF neural networks are separately used to classify and forecast the heavy rainfall and the rainfall situation under the heavy rainfall.Furthermore,the output of the RBF neural networks can be used to revise the PNN neural network model reversely.The reliability and stability of hybrid multi-neural-network models proposed in this paper are verified with the K-folder cross-validation method.The results of the experiments show that the method proposed in this paper can improve the forecast accuracy,recall and achieve good results in the practical application.
滕少华,唐海涛,张巍,刘冬宁,梁路. 混合PNN和RBF多神经网络模型的局域降雨类型识别及雨量预测[J]. 小型微型计算机系统, 2016, 37(11): 2571-2576.
TENG Shao-hua,TANG Hai-tao,ZHANG Wei,LIU Dong-ning,LIANG Lu. Identifying Local Rainfall Type and Forecasting Rainfall Quantity Based on Mixed Multiple PNN and RBF Neural Network Models. Journal of Chinese Computer Systems, 2016, 37(11): 2571-2576.
2016, Vol. 37 
