Day-ahead optimization for railway energy management system

摘要

While rail transportation is a clean and efficient way of transporting freight and persons, a drastic increase in traffic volume is expected in the future. At the same time, a decrease of greenhouse gas emissions is demanded by the European Union (EU) [1]. In response, the EU project MERLIN was brought to life in order to develop an architecture for smart energy management of the European railway systems. This architecture operates on 3 temporal levels, Day Ahead, Minute Ahead and Real Time. The scope of this paper is the methodology of the Day Ahead Optimization algorithm. As first entity in the MERLIN optimization process, the algorithm optimizes the behavior of all participants in the railway system, like trains, buildings and storage systems for the upcoming day. To identify the behavior with the least energy consumption, least energy cost or the lowest power peaks, a problem dependent combination of a genetic algorithm as well as linear and quadratic programming is used. The resulting optimized train, external consumer and storage operating strategies are the basis of further improvement of the system behavior in the Minute Ahead Optimization.