基于混合效应的船舶交通流量预测分析

发布时间：2018-03-25 17:30

本文选题：船舶交通流　切入点：混合效应　出处：《武汉理工大学》2014年硕士论文

【摘要】：船舶交通流量预测是将统计预测方法应用到水运工程技术上的一门新兴学科,其研究对于港口规划建设、管理调度意义重大.目前,对于船舶交通流量的预测方法很多,且有些方法的预测精度高、计算量小,但这些预测方法除具有各自的一般缺点之外,对所有船舶(无论大小)都按一艘进行统计,存在船舶交通流量细节预测精度低等问题,降低了预测结果的可适用性.因此探索新的方法研究船舶交通流量特点并对其进行预测有着极其重要的意义. 本论文主要利用混合效应模型理论,在船舶交通流现状的背景下,将船舶进行分类别研究,根据船舶交通流量的相关因素及船舶个体特征,探索各影响因素之间以及与船舶交通流量之间的混合效应关系,从而建立最大程度反映港口实际交通状况与特性的符合船舶交通流的混合效应模型,利用该模型对船舶交通流量进行预测研究分析,进而为水道或航道的规划设计和船舶通航管理提供基础性依据,并做出科学合理的港口规划. 本文主要工作为以下几个方面： 1.总结影响船舶交通流量的相关因素,并根据到达船舶的个体特征,对船舶进行合理分类,并分析混合效应对于船舶交通流研究的可适用性. 2.根据船舶交通流量影响因素,建立线性混合效应模型,通过拟合比对,验证混合效应模型的可靠性和可行性. 3.建立基于样条的非参数混合效应模型,解决模型计算问题,并将其应用到船舶交通流量与到达规律的研究上. 4.利用天津港船舶交通流观测数据进行实例探究与结果对比分析. 论文详尽地描述了船舶个体特征与交通流量的混合效应关系,结果表明利用混合效应模型,考虑组间差异,区分船舶类型来研究交通流量,验证了模型对于交通流量预测与到达规律研究的可行可靠性与有效性,提高了预测精度,实际应用性很强,并为船舶交通流量细节预测提供了一个好方法.
[Abstract]:Ship traffic volume forecasting is the method applied to the port and waterway engineering technology on a new subject for the study of statistical prediction, port planning and construction, management and scheduling of great significance. At present, the prediction method of ship traffic flow, and some methods of high prediction accuracy, small amount of calculation, but the prediction method has general shortcomings the outside of all ships (regardless of size) according to a statistics, there is low accuracy problem of ship traffic flow prediction for details, reduce the prediction results. So the exploration method of ship traffic flow characteristics and its prediction has extremely important significance.
This paper used the mixed model theory in effect, the ship traffic flow situation under the background of the ship by category research, according to the related factors of vessel traffic flow and ship individual characteristics, mixing effect between the various influencing factors, and explore the relationship between traffic flow, so as to establish the maximum degree reflects the mixed effect model of ship the actual traffic status and traffic flow characteristics of port, forecast analysis of ship traffic flow by using the model, so as to provide basis for channel planning and management of ship navigation, and make scientific and rational planning of ports.
The main work of this paper is the following aspects:
1., summarize the factors that affect the traffic volume of ships, and classify the ships reasonably according to the individual characteristics of the ships, and analyze the applicability of the mixed effects to the research of ship traffic flow.
2. according to the influence factors of the ship traffic flow, a linear mixing effect model is set up, and the reliability and feasibility of the mixed effect model is verified by fitting comparison.
3. a non parametric mixed effect model based on spline is established to solve the problem of model calculation, and it is applied to the study of ship traffic flow and arrival law.
4. use the observation data of the ship traffic flow in Tianjin port to carry out a case study and compare the results with the results.
The detailed description of the relationship between individual characteristics and the mixed traffic flow of the ship, the results show that the mixed effects model, considering the differences between groups, to study the traffic flow between ship types, validation of the model for the study of traffic flow prediction and arrival pattern can be feasible and effective, to improve the prediction accuracy and practical application very strong, and it provides a good method for ship traffic flow details forecast.

【学位授予单位】：武汉理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491.112;O212.1

【参考文献】