基于不同模型的福建省物流需求预测比较研究

Comparative Study on Logistics Demand Forecasting in Fujian Province Based on Different Models

摘要: 区域物流需求研究是制定区域物流规划的基础支撑。利用2013~2021年统计数据，采用灰色关联度模型筛选影响因素，以此建立灰色预测模型及LSTM模型并对福建省物流需求展开预测。研究发现，灰色GM (1, N)模型平均误差为4.96%，LSTM模型预测平均误差为1.52%。预测结果表明LSTM模型较灰色GM (1, N)模型准确性更好，证明LSTM模型用于福建省物流需求预测具备更好的适用性，在此基础上预测2022~2026年结果，为福建省相关部门制定物流规划与政策提供一定的参考意义。

Abstract: The study of regional logistics demand is the basic support for the formulation of regional logistics planning. Based on the statistical data from 2013 to 2021, the grey correlation model was used to screen the influencing factors, so as to establish the grey prediction model and LSTM model and predict the logistics demand in Fujian Province. It is found that the average error of the grey GM (1, N) model is 4.96%, and the average error of the LSTM model is 1.52%. The prediction results show that the LSTM model has better accuracy than the gray GM (1, N) model, which proves that the LSTM model has better applicability in logistics demand forecasting in Fujian Province, and on this basis, the results of 2022~2026 are predicted, which provides a certain reference significance for relevant departments in Fujian Province to formulate logistics planning and policies.