摘要
集装箱船航次配箱量预测对集装箱码头管理和集装箱船配载具有重要意义。首先利用支持向量机(SVM)理论建立非线性回归模型,然后分析影响航次配箱量的因素,利用历史数据作为学习预测的样本,用训练好的回归模型对新的数据进行预测。实际计算结果表明:同BP神经网络预测模型相比,该预测模型具有良好的泛化能力及准确的预测结果。
It is significant to forecast the number of containers for containership voyages. First, the nonlinear regression model based on Support Vector Machine (SVM) is set up, then influencing factors on the number of containers for containership voyages are analyzed. The historical statistic datum is used to train the model and the new value is forecasted. Practical result shows that the method of SVM possesses better regression ability than BP neural network's, which provides a new method to forecast the number of containers for containership voyages.
出处
《中国造船》
EI
CSCD
北大核心
2006年第2期101-107,共7页
Shipbuilding of China
基金
辽宁省教育厅高等学校科研计划基金资助项目(05L091)
高等学校博士学科点专项科研基金资助项目(2000014125)
关键词
船舶、舰船工程
集装箱船
预测
支持向量机
航次配箱量
ship engineering
container ship
forecast
support vector machine
the number of containers for a voyage