基于组合模型的空间信息网络流量预测算法

Traffic prediction algorithm of space information network based on combination model

提出了一种基于经验模式分解(EMD)、遗传优化的广义回归神经网络(GA_GRNN)和自回归移动平均(ARMA)组合模型的空间信息网络流量预测算法。描述了算法的主要理论和具体过程,通过与单一GRNN和ARMA模型的仿真对比,验证了组合模型的有效性。

This paper proposes a novel traffic prediction algorithm for space information network based on the combination models of empirical mode decomposition （EMD）, generalized regression neural network optimized by genetic algorithm （GA_GRNN） and auto regressive moving average （ARMA）. It presents the main theory and the concrete process of the proposed algorithm, and verifies the effectiveness of the combined model by comparing with the simulationof single GRNN and ARMA model.