Enhancing CBTC System Efficiency with DRL and Edge Computing

Rodgers, Alice Angenette (2024) Enhancing CBTC System Efficiency with DRL and Edge Computing. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (500kB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

Communication Based Train Control (CBTC) is becoming increasingly important in optimizing urban transportation systems as they become larger. In this paper we propose a hybrid framework that combines Deep Reinforcement Learning (DRL) and Edge Computing to optimize CBTC systems. Deep Q-Network (DQN) model helps solve dynamic challenges like task offloading, equipment placement, and maintenance scheduling; while, Edge Computing, deployed on AWS Greengrass, reduces latency and computational load by processing the data close to the source. Performance of the framework is evaluated on synthetic data obtained from CBTC system simulations in terms of latency, energy savings, and optimization of cost saving. The model demonstrated that the DQN model outperformed the Q-learning baseline resulting in a 57% reduction in latency, 25% energy savings, and 35% reduction in operational cost. Results were promising for the DQN model and the proposed method has advantages.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Sahni, Vikas UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Cloud computing
Divisions:	School of Computing > Master of Science in Cloud Computing
Depositing User:	Ciara O'Brien
Date Deposited:	16 Jul 2025 11:23
Last Modified:	16 Jul 2025 11:23
URI:	https://norma.ncirl.ie/id/eprint/8146

Actions (login required)

View Item