基于灰色关联与RBF神经网络的港口吞吐量预测

Port Throughput Prediction Based on Grey Correlation and RBF Neural Network

摘要: 在中国的运输体系中，港口扮演着重要的角色。通过建立改进的神经网络算法，运用灰色关联分析，从多个宏观经济指标中提取相关性较强的指标因素来确定RBF神经网络的输入指标。实证结果表明，该模型不仅很好地反映了港口吞吐量与第一产业总产值、第二产业总产值和客运量之间的关系，同时有效提高了预测精度，为决策者提供决策参考。

Abstract: In China’s transportation system, ports play an important role. By establishing the improved neural network algorithm and using the grey correlation analysis, the input index of RBF neural network was determined by extracting the index factors with strong correlation from several macroeconomic indicators. The empirical results show that the model not only well reflects the relationship between port throughput, gross output value of the primary industry, gross output value of the secondary industry and passenger volume, but also effectively improves the prediction accuracy and provides reference for decision makers.