摘要
港口吞吐量的预测是港口决策的重要依据，然而影响港口吞吐量的要素较多，且各要素之间存在着复杂的非线性关系，使常用的预测方法难以取得好的预测效果。文章试图运用BP人工神经网络的方法，采集1999～2012年的实际数据样本，以钦州港为例进行实证分析，构建三层BP神经网络，预测2013年钦州港港口吞吐量。
The forecast of port throt, ghput is an important basis for decision-making of port, however there many factors influenet the port throughput, and the complex nonlinear relationship existed between each element and each element, which make the usual prediction methods obtain good prediction effect difficulty. Applying the method of artificial neural network, through the ac- tual data of 1999-2012 samples to do the empirical analysis, constnlct the three layers BP neural network, and predict the Qinzhou port throughput in 2013.
出处
《物流科技》
2014年第1期56-59,共4页
Logistics Sci-Tech
基金
钦州学院科研资助项目
项目编号:2012XJKY-30B
关键词
系统工程
港口吞吐量
BP神经网络
预测模型
钦州港
systems engineering： port throughput
BP neural network： prediction model： Qinzhou port
