摘要
快递业作为现代经济体系的关键基础设施,在维护产业链供应链稳定、驱动居民消费升级及保障国内国际双循环系统效能等方面发挥着战略性作用。文章通过建立GM(1,1)-BP神经网络组合预测模型,创新性地将灰色系统的时间序列特征提取能力与神经网络的非线性映射特性相结合,采用动态权重分配机制实现模型优势互补。以湖南省快递业发展为例,构建了包含5个维度16项指标的快递需求预测指标体系,运用GM(1,1)-BP神经网络组合模型进行快递需求预测。结果表明,组合预测模型有效提升了预测的准确性,具有显著统计学意义。最后,根据预测结果提出了湖南省快递业高质量发展的对策建议。
As a critical infrastructure in the modern economic system,the express delivery industry plays a strategic role in stabilizing industrial and supply chains,driving the upgrading of residential consumption,and ensuring the efficiency of domestic and international dual-circulation systems.This study innovatively proposes a combined forecasting model integrating GM1,1 and BP neural networks,which leverages the time-series feature extraction capability of the grey system and the nonlinear mapping characteristics of neural networks.A dynamic weight allocation mechanism is introduced to synergize the advantages of both models.Taking the development of Hunan Province's express delivery industry as a case study,a predictive indicator system comprising sisteen metrics across five dimensions is constructed.Empirical results demonstrate that the hybrid model significantly improves forecasting accuracy,achieving statistically meaningful outcomes.Based on the predictions,targeted recommendations are proposed to support the high-quality development of Hunan's express delivery sector.
作者
李青松
王奕博
秦荣怀
LI Qingsong;WANG Yibo;QIN Ronghuai(School of Economics and Management,Central South University of Forestry and Technology,Changsha 410004,China)
出处
《物流科技》
2025年第11期12-17,共6页
Logistics Sci Tech
基金
国家社会科学基金一般项目(22BGL114)
教育部人文社会科学研究青年基金项目(24YJCZH141)。
关键词
快递需求
GM(1
1)
BP神经网络
组合预测
express delivery demand
GM(1,1)
BP neural network
combination forecasting
