Research on Spatio-Temporal Prediction and Rebalancing Optimization for Bike-Sharing Supply-Demand Imbalance

Abstract

Bike-sharing systems have emerged as a pivotal component of urban sustainable transportation, yet they frequently face challenges related to supply-demand imbalances across spatial and temporal dimensions. This study aims to address these inefficiencies by integrating spatio-temporal prediction with rebalancing optimization strategies. The proposed methodology employs a hybrid deep learning model combining convolutional neural networks (CNN) and long short-term memory (LSTM) networks to forecast short-term bike demand and supply patterns across stations. Subsequently, a rebalancing optimization model is formulated using integer programming to minimize operational costs while ensuring service availability. Experimental results on real-world bike-sharing data demonstrate that the hybrid prediction model achieves superior accuracy in capturing spatio-temporal dependencies, and the optimization approach significantly reduces imbalance rates by 18.7% compared to baseline methods. The findings underscore the importance of data-driven decision-making for enhancing system efficiency and user satisfaction in shared mobility services. This research provides actionable insights for urban planners and bike-sharing operators to improve resource allocation and operational sustainability.