A hybrid optimization algorithm for energy efficient train operation

In this manuscript, an energy-efficient train operation between successive stations is studied. Cruising and coasting, two basic motion phases of train, should be taken into consideration in order to decrease energy consumption. Determining the optimal switching points from one motion phase into another is key in the energy saving. It is shown that, genetic algorithm and simulated annealing, when employed in a hybrid algorithm, complement each other in finding such switching points. For a performance verification of the hybrid optimization approach, multiple test tracks with different lengths are considered. Also certain real life constraints are taken into account such as punctuality and maximum speed limit. Obtained results are compared to the single genetic algorithm and it is shown that the hybrid algorithm built as a cascade combination of genetic algorithm and simulated annealing can reach optimum solution with better accuracy and lower time consumption.