关键词: 灰色系统 流量预测 GM（1 1）模型 径向基神经网络

Abstract:

According to the idea that neural networks can effectively modify grey prediction models, the paper proposes a combination prediction model based on r  adial basis function neural networks. By collecting the network switching data of the campus, the paper builds a grey GM（1,1） model based on analysingg the netwok traffic time sequence characteristics,and corrects the remain differences of the model based on adopting the RBF neural network. The experi  mental results show that the novel model is more effective than the single grey model.

Key words: grey system, network traffic prediction, GM（1,1） model, RBF neural network