Analysis of Logistics Demand Distribution in Beijing Based on Kernel Density Estimation

Abstract

Identifying regions with high demand for urban logistics can lay the foundation for studies related to urban logistics, such as the choice of distribution routes and the layout of logistics facilities. This paper presents kernel density estimation (KDE) method to study the issues related to the spatial distribution of urban logistics demand. The paper also applies this method to the study of logistics demand distribution in Beijing. First, it selects the Gaussian kernel as the kernel function to obtain the heat map of probability density distribution of logistics demand in Beijing. Secondly, this paper finds four regions with higher distribution density throughout the year. Additionally, the paper discusses the characteristics of monthly logistics distribution in these regions and their causes. Finally, it uses the rank sum test to analyze whether the differences in demand distribution of the four regions are significant.