混合径向基神经网络对货运量的预测

为了定量预测多个外部因素影响下的货运量,建立了混合径向基神经网络模型.该模型以径向基神经网络为模型主体,并结合二阶振荡粒子群优化算法和灰色预测方法构成混合预测模型.该神经网络模型的参数设置更加简便,收敛速度更快.实例预测得到的结果相比较其他预测方法绝对误差值更小,误差变化范围更加稳定,证实了该神经网络模型的有效性,表明了其在多因素影响下的货运量预测中具有很好的适用性.

Mixed Radial Basis Function Neural Network in Freight Volume Forecasting

In order to quantitatively predict the freight volume with the effects of external factors,a mixed radial basis function neural network model was built,which was based on the radial basis function neural network as the model of main part,and combined of two-order oscillating particle swarm optimization algorithm and gray forecast method.This mixed neural network model had the advantages of simpler parameters settings and faster convergence speed than other traditional methods.The calculated result of this model had smaller absolute errors and more stable error variation range compared with other prediction methods.The results demonstrated the validity of the mixed neural network model and indicated the good application of freight volume forecasting under the influence of multiple factors.