Study on Demand Forecasting and Scheduling Routes of Shared Bicycles

This paper endeavors to address the pressing issue of resource wastage in shared bicycles by proposing an innovative approach to optimize their utilization and cater to the demands of urban residents. The proposed solution involves devising an efficient vehicle dispatch roadmap based on predictive demand modeling. Leveraging the open-source Beijing shared bicycle dataset, the research analyzes the spatio-temporal correlations within order data. The Temporal Graph Convolutional Network (T-GCN) is selected as the predictive model to forecast shared bicycle demand. Subsequently, the Genetic Algorithm (GA) is employed to determine an optimal dispatch route, thereby significantly improving the overall utilization rate of shared bicycles.