节点文献
新建港口吞吐量预测研究
Research on the Forecast of New Port Throughput
【作者】 李丹;
【导师】 施先亮;
【作者基本信息】 北京交通大学 , 物流工程(专业学位), 2018, 硕士
【副题名】以小漠港为例
【摘要】 深圳盐田港集团为贯彻国家“一带一路”和泛珠三角区域合作战略,投资建设了深汕特合区小漠国际港及其临港工业园,其目标将小漠港打造成为国际物流港,承接深圳的产业转移和转型升级的同时发挥“以港兴城”的作用。小漠港进行港口规划时,过大、过小的规模的港口建设都会造成资源的浪费,因此,需要对港口货物吞吐量进行合理预测,才能为港口布局、经营策略、基本设施投资规模等方面提供决策依据。由于小漠港为新建港口,缺乏历史吞吐量数据,但现有文献多数是在港口的历史吞吐量基础上进行的,并且只集成各类预测模型进行研究。本文提出集成预测的方法,包括集成多维信息、多种方法、各类模型来估算小漠港的吞吐量。从而丰富了港口吞吐量预测的思路和方法,为企业建立经营、投资计划提供依据,同时可借鉴于其他新建港口、新建物流园区等的吞吐量预测。首先,在小漠港的腹地范围内分析其吞吐量影响因素,包括小漠港自身的港口建设规划,其腹地的经济发展情况,周边港口与小漠港间的竞合关系,以及宏观的外部环境四维因素。由此分析小漠港的货物吞吐量由自身腹地带来的货物量以及从周边其他港口吸引的货物量两部分组成。接着,类比其他20个沿海港口,集成采用定性分析、相关性分析、灰色关联度分析三次筛选提取静态物流量的影响因素,并利用主成分回归拟合港口吞吐量预测模型;并基于GM(1,1)模型预测港口静态吞吐量关键影响因素,结合预测模型输出小漠港静态吞吐量。再者,结合港口间的竞合关系,通过熵权TOPSIS法量化港口的竞合度,以此融合引力模型,计算小漠港对不同港口的吸引力系数;并基于各港口的历史吞吐量利用Holt指数平滑模型预测各自未来的吞吐量;结合各港口所在圈层权重,加权求和得出小漠港从周边港口吸引的动态物流量。最终预测出新建小漠港2020年的总吞吐量,经过与深汕特合区组委会各专家的交流,该预测值较为符合实际,并通过其他港口验证了本文集成预测思路,结果表明具有一定的可靠性。论文对新建港口的吞吐量预测以及港口物流规划提供了科学的理论依据和决策支持。
【Abstract】 In order to implement the national "One Belt,One Road" and Pan-Pearl River Delta regional cooperation strategy,Shenzhen Yantian Port Group invested in the construction of the Xiaomo International Port and its port Industrial Park in the ShenShan Special Zone.The goal is to build Xiaomo Port into an international logistics port,undertaking Shenzhen’s industrial transformation and upgrading,and realizing the policy of "Port Promote the City" as well.When the port planning is carried out in Xiaomo Port,the construction of ports with too large or too small scale will cause waste of resources.Therefore,it is necessary to make a reasonable forecast of the port cargo throughput so that it can provide decision-making basis for port layout,business strategy,investment in basic facilities,etc.Xiaomo Port is a newly-built port and lacks historical throughput data,but most of existing literatures are based on the historical throughput of the port,and only integrate various types of prediction models for research,which means they don’t match Xiaomo Port.This paper proposes an integrated forecasting method that includes integrating multidimensional information,multiple methods,and various models to estimate the throughput of Xiaomo Harbor.This enriches the ideas and methods for port throughput forecasting,provides a basis for companies to establish operations and investment plans,and can draw on the throughput forecasts of other new ports and new logistics parks.Firstly,the factors affecting the throughput were integrated and analyzed in the hinterland of Xiaomo Port,including the port construction plan of Xiaoxiao Port,the economic development of its hinterland,the cooperative relationship among surrounding ports and Xiaomei Port,and macroscopical external environment.According to the analysis,the cargo throughput of Xiaomo Port consists of the volume of cargo brought by its own hinterland and the amount of cargo that is attracted from other surrounding ports.Then,analogous to the other 20 coastal ports,the paper integrates qualitative analysis,correlation analysis,and grey correlation analysis to extract the influencing factors of static flow,and uses the principal component regression to fit the port throughput prediction model.Based on GM(1,1)model,the key influencing factors of the port’s static throughput are predicted,and combined as the forecast model to output the static throughput of Xiaomo Harbor.In addition,this paper combines the actual relationship between ports to quantify the degree of convergence of the port through the entropy weight TOPSIS method,and then fuses the gravity model to calculate the attraction coefficient of the Xiaomo Port to different ports.Based on the historical throughput of each port,this paper uses the exponential smoothing model to predict their future throughput,combines with the division results result of ShenShan Special Zone traffic planning to the circle of Xiaomo Port hinterland.And also,a weighted summation is used to draw the dynamic logistics of the Xiaomo Port.It is predicted that the total throughput of the newly-built Xiaomo Port will reach 49.5 million tons in 2020.After exchanges with the experts of the Organizing Committee of the ShenShan Special Zone,the predicted value is more realistic and has a certain reliability.The paper provides scientific theoretical basis and decision support for the throughput prediction of new ports and port logistics planning.