Campus Units

Civil, Construction and Environmental Engineering

Document Type


Publication Version

Accepted Manuscript

Publication Date


Journal or Book Title

IEEE Transactions on Intelligent Transportation Systems




This paper presents a real-time dispatching model for electric autonomous vehicle (EAV) taxis that combines mathematical programming and machine learning. The EAV taxi dispatching problem is formulated and solved as an integer linear program that maximizes the total reward for serving customers. The optimal dispatch solutions are generated by simulating electric autonomous taxis that are dispatched by the optimization model. The artificial-neural-network-(ANN)-based model was trained using the optimization model's dispatch solutions to learn the optimal dispatch strategies. Although the dispatch decisions made by the ANN-based model are not optimal, the system's performance is very close to the optimization dispatch model in terms of customer service and taxis' operational efficiency. In addition, the ANN-based dispatch model runs much faster. By comparing with current taxis, it was found that the EAV taxis dispatched by our ANN-based model can improve operational efficiency by reducing empty travel distance. EAV taxis can also reduce fleet size by 15% while maintaining a comparable level of service with the current taxi fleet.

Research Focus Area

Transportation Engineering


This is a manuscript of an article published as Hu, Liang, and Jing Dong. "An Artificial-Neural-Network-Based Model for Real-Time Dispatching of Electric Autonomous Taxis." IEEE Transactions on Intelligent Transportation Systems (2020). DOI: 10.1109/TITS.2020.3029141. Posted with permission.


© 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Copyright Owner




File Format


Published Version