一种打车需求预测方法及系统
By constructing a ride-hailing demand prediction model using graph neural network and generative neural network modules, and combining autocorrelation and inter-segment correlation, the problem of traditional systems being unable to accurately predict ride-hailing demand is solved, achieving efficient demand prediction and resource scheduling, and improving the efficiency of taxi services and customer satisfaction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- UNIV OF SCI & TECH BEIJING
- Filing Date
- 2024-06-18
- Publication Date
- 2026-07-17
AI Technical Summary
Traditional ride-hailing dispatch systems cannot accurately predict and respond to real-time changes in ride-hailing demand, resulting in high customer delay rates, insufficient utilization of taxi resources, and negatively impacting customer experience.
A ride-hailing demand prediction model is constructed using a graph neural network module, a self-attention mechanism module, and a generative neural network module. Through autocorrelation and partial autocorrelation analysis, and by combining the correlation values between road segments to construct a graph structure, the model captures the spatiotemporal characteristics of ride-hailing demand and performs demand prediction.
It enables accurate prediction of future ride-hailing demand, reduces customer waiting time, improves taxi operating efficiency, and optimizes the transportation system.
Smart Images
Figure CN118674212B_ABST