灰色线性回归模型在港口吞吐量预测中的应用_陈秀瑛.pdfVIP

2010 年5 月 水运工程 May 2010 第5 期 总第441 期 Port Waterway Engineering No. 5 Serial No. 441 灰色线性回归模型在港口吞吐量预测中的应用 陈秀瑛，古 浩 （南京水利科学研究院勘测设计院，江苏南京210029） 摘要：利用灰色模型和线性回归模型，分别基于最小二乘及有效度原理建立了两种灰色线性回归模型，并将其用于镇江 港吞吐量预测，可得结论：基于最小二乘原理建立的灰色线性回归模型，预测精度优于单一的灰色模型和线性回归模型，其 应作为镇江港吞吐量预测的最优模型；基于有效度原理建立的灰色线性回归模型预测精度尽管低于线性回归模型，但由于其 本质上为组合模型，能综合多种信息，用于港口吞吐量预测时也具有一定的实用性。 关键词：吞吐量预测；灰色线性回归；最小二乘原理；有效度 + 中图分类号：U 652.1 4 文献标志码：A 文章编号：1002-4972 （2010）05-0089-04 Application of gray linear regression model for forecast of port throughput CHEN Xiu-ying, GU Hao (Survey and Design Institute under Nanjing Hydraulic Research Institute, Nanjing 210029, China) Abstract: By using the gray model and regression model, two kinds of gray linear regression models are set up on the basis of the least square and effectiveness indicator principles respectively. Applying these models to forecast the throughput of Zhenjiang port, the following conclusions can be drawn: The forecast precision of gray linear regression model based on the least square principle is better than that of the simple gray model and linear regression model. This model should be taken as the most superior model in forecasting Zhenjiang ports throughput. The forecast precision of gray linear regression model based on the effectiveness indicator principle is relatively low. But as a combined model, it can synthesize the multiple messages, so it has certain practicability when used to forecast port throughput. Key words: forecast throughput; gray linear regression; least square; effectiveness indicator 港口吞吐量预测是港口规划及其可行性研究