基于GRNN神经网络的长江干线港口集装箱吞吐量预测  

Container Turnover Forecast of Link Ports alongside the Yangtze River Based on GRNN

作　　者：赵亚鹏[1] 丁以中[1] 

机构地区：[1]上海海事大学,上海200513

出　　处：《中国航海》2006年第4期90-91,100,共3页Navigation of China

基　　金：国家自然科学基金项目(70371012);上海市重点学科建设项目资助(T0602)

摘　　要：在长江集装箱货源市场调查分析的基础上,根据长江干线集装箱发展形势,结合长江干线港口发展情况及历年数据调查,考虑影响港口吞吐量预测的复杂因素,运用GRNN神经网络的优点,构建预测模型。预测结果证明,该模型在应用中是有效的,且当样本数据短缺时,预测效果也较佳。On the basis of investigation and analysis of the container supply market alongside the Yangtze River, a forecast model is constructed by taking the advantages of GRNN, in which the development situations of link ports and container, survey data for years, and complicated factors about port throughput forecast are considered. Forecast results have proved that the model is effective in the forecast of link port container tumover and even in the case of sample data shortages.

关 键 词：水路运输 集装箱 吞吐量 预测 

分 类 号：U695.22[交通运输工程—港口、海岸及近海工程]