Techno-Economic Assessment of an AI-Based Railway Station Energy Management System

Previous studies have demonstrated the technical effectiveness of artificial intelligence (AI)-based energy management systems (EMS) for railway stations using an artificial neural network (ANN) and a Deep Q-Network (DQN) to optimize energy operation while utilizing photovoltaic (PV) and regenerative braking energy. This study evaluates the techno-economic feasibility of the AI-based EMS by considering capital expenditure (CAPEX), operating expenditure (OPEX), and annual economic benefits. Economic performance is assessed using the payback period, net present value (NPV), internal rate of return (IRR), benefit-cost ratio (B/C), and sensitivity analyses of ESS price, electricity-price escalation, and regenerative energy utilization. The results indicate that the proposed EMS is economically feasible and provides a practical basis for evaluating AI-based railway station energy management systems.