基于多模型组合的物流需求预测分析——以武汉市为例  

A Multi-Model Based Combination Prediction and Analysis of Logistics Demand: Taking Wuhan as Example

作　　者：武亚鹏 李慧颖 李婷 刘淮源 WU Yapeng;LI Huiying;LI Ting;LIU Huaiyuan(Central South Survey Design Institute Group Co.,Ltd.,Wuhan 430073,China)

机构地区：[1]中南勘察设计院集团有限公司,湖北武汉430073

出　　处：《物流技术》2022年第6期60-63,共4页Logistics Technology

摘　　要：采用有效度法对线性回归、ARIMA模型以及灰色预测模型进行线性组合,经验证,其最大误差为5.56%,平均误差为2.94%,处于合理范围。运用组合模型对武汉市2021-2030年物流需求(货运量)进行预测分析,经计算至2030年武汉市年货运总量将达到140 000万吨以上,接近上海等一线城市现有物流需求水平。面对巨大的物流需求规模,武汉市在“十四五”期间,有必要制定物流业中长期发展规划,以指导现代化物流体系的建立,推动物流业降本增效。In this paper, to forecast and analyze the logistics demand(as represented here by freight volume) in Wuhan from 2021 to2030, we linearly combined the linear regression, ARIMA and gray forecast models based on validity, which, as we verified, was 5.56% in maximum error and 2.94% in average error, both in a reasonable range of accuracy. Next, we used the combination model to forecast the logistics demand of Wuhan from 2021 to 2030 and calculated that by 2030, the annual freight volume of Wuhan would exceed 1.4 trillion tons, approximating the current logistics demand level in first-tier cities like Shanghai. Finally, we proposed that in face of logistics demand of such magnitude, it is imperative, during the "14th Five-Year Plan" period, for Wuhan to formulate a medium and long-term development plan for the logistics industry to guide the establishment of a modern logistics system and promote the cost-efficient development of the logistics industry.

关 键 词：物流需求 模型组合 灰色预测模型 线性回归 时间序列法 武汉市 

分 类 号：F259.27[经济管理—国民经济] F224.0