基于时间序列模型的全国港口货物吞吐量的预测分析

Predictive Analysis of Cargo Throughput of Ports Nationwide Based on Time Series Model

摘要: 为把握全国港口货物吞吐量的变化趋势，为相关决策提供支撑，本文以2019年1月~2024年12月的全国港口货物吞吐量月度数据为基础，运用时间序列分析方法构建ARIMA模型进行拟合与预测。经数据预处理、平稳性检验、模型构建与优化后，确定ARIMA(0,(1,12),1)为最优模型，该模型拟合效果良好且通过相关检验。预测结果显示，2025年全国港口货物吞吐量将呈稳步增长态势，具有显著的季节性特征。

Abstract: To understand the changing trends of cargo throughput at ports nationwide and provide support for relevant decision-making, this paper uses monthly data on national port cargo throughput from January 2019 to December 2024 as a basis, and employs time series analysis to construct a model for fitting and prediction. After data preprocessing, stationarity testing, model construction, and optimization, the optimal model was determined, demonstrating good fit and passing relevant tests. The prediction results show that national port cargo throughput will exhibit a steady growth trend in 2025, with significant seasonal characteristics.