为了能够给区域物流发展政策的制定、物流基础设施建设规模的确定、物流市场态势的分 析等提供定量的物流需求规模数据，建立科学合理的预测模型显得尤为重要。首先，研究区域物 流与区域经济的关系；其次，从货运量、货运周转量两个指标中选取货运周转量来表征区域物流 需求规模；最后，从区域经济指标中选取第一产业总产值、第二产业总产值、第三产业总产值、 社会消费品零售总额、固定资产投资额、进出口额等指标作为影响因素，借助SPSS 统计分析软 件，以云南省统计数据为基础，建立基于主成分回归方法的区域物流需求预测模型。研究结果表 明，该模型在对云南省物流需求规模进行预测时，模型平均相对误差小于4%，预测精度高，可 以作为中短期物流需求预测的工具。

In order to provide the quantitative scale data of the regional logistics demand for making the regional logistics development policies, determining the logistics infrastructure construction scale, analyzing logistics market situation and so on, the establishment of scientific and reasonable prediction mod⁃ el is particularly important. Firstly, the relationship between regional logistics and regional economy was studied. Secondly, the freight turnover quantity was selected from the freight quantity and the freight turnover quantity to characterize the regional logistics demand scale. Finally, primary industry value,secondary industry value, tertiary industry value, total retail sales of consumer goods，fixed asset invest⁃ ment volume, value of export and import etc., were selected from regional economy indexes to be as influencing factors. Based on statistical data of Yunnan Province, using SPSS statistical analysis software, the region logistics demand forecasting model based on principal component regression method was established. The research confirmed that the model for forecasting the logistics demand scale of Yunnan Province, the average relative error of the model is less than 4%, the model has higher prediction accuracy and can be used as medium and short term logistics demand forecasting tool.