Author(s)

Hanxin Chen

Corresponding Author

Hanxin Chen

Abstract

With the rise of e-commerce platforms and the popularity of online shopping, the express delivery industry is experiencing rapid demand growth. Accurate demand forecasting has become crucial for optimizing logistics networks, reducing costs, and enhancing service quality. This paper proposes a comprehensive framework for express demand forecasting based on the Entropy Weight-TOPSIS and ARIMA models. First, the Entropy Weight-TOPSIS model is used to evaluate the importance of cities in the express logistics network. Six indicators, including the number of supply cities, the number of receiving cities, average shipment volume, average receipt volume, shipment change rate, and receipt change rate, are selected to rank 24 cities comprehensively. The results show that cities L, G, V, W, and B have the highest importance in the express logistics network. Subsequently, the ARIMA model is employed for time-series analysis and forecasting of express demand. Through time-series modeling of the total express volume, combined with ADF tests and autocorrelation analysis, an ARIMA(0,1,2) model is established to predict express transportation volumes for the next two days. The model exhibits good forecasting performance with a fit of R² = 0.837. The study demonstrates that the integrated approach combining the Entropy Weight-TOPSIS and ARIMA models can effectively evaluate the importance of cities in the express logistics network and provide accurate references for express demand forecasting, offering strong support for logistics planning and resource allocation by express companies.

Keywords

Express demand forecasting, Entropy Weight-TOPSIS model, ARIMA model, Logistics network, City importance evaluation, Time-series analysis