Analyzing and Forecasting Container Throughput With a Hybrid Decomposition‐Reconstruction‐Ensemble Method: A Study of Two China Ports

• DOI: 10.1002/for.3253
• Publication Date: 2025-01-06T07:59:52Z
• Authors: Yi Xiao; Sheng Wu; Chen He; Yi Hu

ABSTRACT Accurate container throughput forecasting is critical for enhancing port efficiency and ensuring global trade stability, particularly in the face of economic uncertainties, geopolitical tensions, supply chain disruptions. Existing methods often struggle to model nonlinear, nonstationary, noise‐laden characteristics data, creating a clear gap ability provide reliable predictions. To address this, we propose novel hybrid model, VMD‐ISE‐TCNT, designed tackle these challenges. The employs variational mode decomposition (VMD) decompose time series into intrinsic modes, with an improved signal energy (ISE) criterion automating selection optimal numbers. These modes are categorized low‐ high‐frequency components forecasted separately using temporal convolutional networks (TCNs), leveraging their strength capturing multiscale dependencies. Theil UII‐S loss function integrated enhance robustness by prioritizing proportional accuracy reducing outlier sensitivity. Empirical evaluations 24 years data from China's two largest ports—Shanghai Shenzhen—demonstrate superior performance VMD‐ISE‐TCNT compared traditional benchmarks. By addressing frequency‐specific patterns key processes, this provides scalable interpretable solution advancing operations resilience trade.