基于神经网络预测的长江引航量周期变化模型
Periodic variation model of pilotage volume in the Yangtze River based on neural network prediction
卢萍 1石文宝2
作者信息
- 1. 上海海事大学商船学院,上海 201306;长江引航中心,江苏太仓 215400
- 2. 上海海事大学商船学院,上海 201306
- 折叠
摘要
为研究引航量周期变化的问题,提出通过神经网络预测的方法,对长江引航量长周期和短周期变化进行研究.在长江引航量动态分析的基础上,借助BP(Back Propagation)神经网络经验公式和反复训练的方法,构建长江引航量的长周期和短周期变化模型,通过回归拟合度数值比较,验证BP神经网络预测的准确性.从长江引航量变化规律的研究发现,长江引航量长周期变化规律和波罗的海干散货指数(IBD)关联性强,短周期变化规律受国内市场和长江航运等多变量影响.引航量的预测研究为引航机构的发展规划和资源布局提供合理的数据、决策支持,使港口能够高速、安全的发展;周期变化的规律将应用于疫情防控下航运影响分析,为精准疫情防控和引航标准与规范服务提供科学参考.
Abstract
In order to study the problem of cyclic variation of pilotage volume,this paper proposes to study the long-cycle and short-cycle variation of the Yangtze River pilotage volume by the method of neural network prediction.Based on the dynamic analysis of the pilotage volume of the Yangtze River,this paper constructs the long-period and short-period variation models of the pilotage volume of the Yangtze River by means of the BP(Back Propagation)neural network empirical formula and repeated training method,and verifies the accuracy of the prediction of the BP neural network through the comparison of the regression fitting degree.The study on the variation rule of the Yangtze River pilotage volume shows that the long-period variation rule of the Yangtze River pilotage volume has a strong correlation with the IBD(Baltic Dry Index),while the short-period variation rule is affected by the domestic market and the Yangtze River shipping.The forecast study of pilotage volume provides reasonable data and decision support for the development planning and resource layout of pilotage organizations,so as to enable the port to develop safely in a high speed.The cycle change pattern will be applied to analyze the impact of shipping under the epidemic,which will provide scientific reference for precise epidemic prevention and control,and for pilotage standard and normative services.
关键词
水路运输/变化模型/神经网络/长江引航量/回归拟合Key words
waterway transportation/variation model/neural network/the Yangtze River pilotage/regression fitting引用本文复制引用
基金项目
上海高水平地方高校创新团队()
2022年度科委地方院校能力建设计划项目()
出版年
2024
NETL
NSTL
万方数据