Abstract: Aimed at the precision of country economy forecasting，Radial Basis Function（RBF） neural network is applied in the field of the country economic forecasting.Double-layers neural network architecture is constructed so that framework and parameters are determined by learning and training.Ultimately，RBF neural network forecasting model is established.Compared with the BP neural network model by forecasting，the results show that the RBFNN（Radial Basis Function Neural Network）’s forecasting precision and generalization ability is more superior to BP’s.Besides，a problem-solving idea and method for the economic forecasting work is provided.

Key words: neural network, Radial Basis Function（RBF）, Gross Domestic Product（GDP）, forecasting model

摘要： 为了建立国民经济生产总值（GDP）神经网络预测模型，构造了双层网络结构的基于径向基网络，通过学习训练，确定径向基数神经网络参数和结构。仿真结果表明，生成的径向基函数模型应用于国民经济预测比BP神经网络模型具有更高的预测精度和良好的泛化能力。

关键词: 神经网络, 径向基函数, 国内生产总值, 预测模型