Energy management system for DC railway smart grid based on substation power forecast and energy storage system optimization

This paper presents a day-ahead energy management strategy for a DC smart railway grid integrating a photovoltaic (PV) power generator and energy storage systems (ESS) at the traction substation (TSS). The proposed energy management method relies on the forecasted power consumption profile of the TSS for the upcoming day and a non-linear optimization approach to manage the ESS in an economically efficient way while respecting physical constraints and maintaining power balance within the power system. An adequate forecast method is used to address the unique characteristics of TSS power consumption, particularly the occurrence of multiple short-duration peak loads during the day. The non-linear optimization approach accounts for both the system’s behavior and the inherent non-linearity of the problem. The forecasted TSS consumption profile is combined with a PV production profile as input to the optimization algorithm, which determines the optimal ESS schedule for the next 24 h. The impact of forecast accuracy on optimization outcomes is evaluated using key performance indicators. To demonstrate the superiority of the proposed optimization method in scheduling the ESS, its results are compared against those of a basic baseline case.