Akademik Veri Yönetim Sistemi
8th IEEE International Conference on Intelligent Systems (IS), Sofija, Bulgaristan, 4 - 06 Eylül 2016, ss.692-697
Increasing demand for energy and environmental concerns have increased importance of energy-efficient transportation. In this manuscript, an energy-efficient train operation based on finding optimal speed profile is studied. As a nature-inspired metaheuristic approach, Firefly algorithm is employed to find optimal switching points of the train control signal. In problem formulation, energy consumption is taken as major part of objective function and travel time is being included as penalty factor. In order to verify the obtained results, a simulation is performed. Besides firefly algorithm, genetic algorithm is also used to compare results. Two algorithms are simulated on test track with various grade profiles for several times. Both of them considered four phases (maximum acceleration, cruising, coasting and braking) of train motion. Furthermore with the help of relaxing boundary conditions, algorithms are able to organize motion phases excluding cruising phase. Simulation results demonstrated that, compared to the GA, FA provides more accurate and persistent solutions. In addition, it can be converged to solution in small iterations, so it is compatible for using in real time problem solving.