Energy-efficient train operation strategy for human-driven trains and a real-world test on the Istanbul Subway System

Abstract

Human-operated train driving is often instinctive and inefficient in terms of energy use, as drivers control speed using the acceleration lever. In this study, unlike existing studies, the use of the acceleration lever was optimized using the Genetic Simulated Annealing (GSA) algorithm by considering the speed stages. This has been tested along a part of the Istanbul Metro line and shown to be easily applicable in practice. Two case studies, including simulation and real-world tests, were conducted with three scenarios: straightforward strategy, improved strategy, and the recommended (optimal usage of acceleration lever) strategy. The simulation results revealed that the recommended method achieves energy savings of over 36% compared to the straightforward method, and over 31% based on real-world tests. The findings suggest that this optimization approach is easily implementable in real-world applications.