基于灰色组合模型的福建省港口集装箱吞吐量预测研究

发布时间：2018-10-17 11:43

【摘要】：集装箱物流作为港口物流中的重要组成部分，是衡量港口在国际贸易中地位的标志之一，集装箱吞吐量的预测也是港口发展战略研究的重要内容。福建省沿海港口作为全国港口的重要组成部分，在全国综合交通体系和对台贸易运输中有着重要的地位。随着海峡西岸经济区建设逐步推进和两岸“三通”的实现，福建省港口发展将面临着巨大的机遇。目前，由于福建省港口腹地经济迅速发展、宏观政策以及国内外宏观经济等各种因素的影响，对港口集装箱吞吐量预测得到的结果和实际现状还存在一定的差距。本文正是在这种背景下提出对福建省港口集装箱吞吐量进行更准确的预测，，为集装箱运输规划和建设提供理论依据。首先，综述了用于港口集装箱吞吐量预测的理论和方法，阐述灰色模型及其组合模型在港口集装箱吞吐量预测中的应用现状，针对港口集装箱吞吐量影响因素的不确定性及其复杂的关系，在预测前期充分分析港口集装箱吞吐量数据规律及特点，建立港口集装箱吞吐量预测方法体系。其次，分析研究GM(1，1)模型在建模过程中存在的问题及其改进方法，提出了一种粒子群算法优化GM(1，1)模型背景值和初值方法，然后将线性回归模型和粒子群算法优化灰色GM(1，1)模型结合，建立了灰色线性组合预测模型。将该组合模型应用到福建省港口集装箱吞吐量的预测中，并与单项预测模型的预测结果进行比较，结果表明本文提出的组合模型的预测结果较为合理。最后，结合预测结果对福建省发展集装箱物流提出了一些建议。
[Abstract]:Container logistics, as an important part of port logistics, is one of the symbols to measure the status of ports in international trade. The prediction of container throughput is also an important part of port development strategy research. As an important part of the national ports, the coastal ports of Fujian Province play an important role in the national comprehensive transportation system and the trade and transportation to Taiwan. With the construction of the economic zone on the west side of the Taiwan Strait and the realization of the "three links" between the two sides, the port development of Fujian Province will face great opportunities. At present, due to the rapid development of port hinterland economy in Fujian Province and the influence of macroeconomic policies and domestic and foreign macroeconomic factors, there is still a certain gap between the results of port container throughput prediction and the actual situation. It is under this background that this paper puts forward a more accurate prediction of container throughput in Fujian port and provides a theoretical basis for container transportation planning and construction. Firstly, the theory and method of container throughput prediction are summarized, and the application status of grey model and its combination model in port container throughput prediction is described. In view of the uncertainty of the influencing factors of port container throughput and its complex relationship, the law and characteristics of port container throughput data are fully analyzed in the early stage of prediction, and the forecasting method system of port container throughput is established. Secondly, the problems existing in the modeling of GM (1K1) model and its improved method are analyzed. A particle swarm optimization method is proposed to optimize the background value and initial value of the GM (1K1) model. Then the linear regression model and the PSO optimization grey GM (1K1) model are combined. The grey linear combination prediction model is established. The combined model is applied to the forecasting of container throughput in Fujian port, and compared with the forecasting result of single forecasting model. The results show that the forecasting result of the combined model is reasonable. Finally, some suggestions are put forward for Fujian Province to develop container logistics.
【学位授予单位】：福建农林大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F552.7;F224;U695.22

【参考文献】