Multi-objective train operation optimization based on fusion distance and preference information

In this paper, the complex multi-objective optimization problem is investigated for the train operation process. To sovle the multi-objective optimization problem, the fruit fly optimization algorithm (FOA) is adopted. Fistly, a new method of distance fusion, which fuses the Mahalanobis distance and the Euclidean distance, is proposed to solve the problem that the classical Mahalanobis distance and Euclidean distance can't effectively calculate the actual distance between the individual and the extremum individual solution set caused by the correlation ambiguity of characteristic variables. Then, to maintain the population diversity, the preference information is adopted so that the optimal solution provided by FOA can significantly move to the desired region. Hence, both problems can be overcome by the improved FOA based on fusion distance and preference information. Finally, simulation results are carried out to show the efficacy of the proposed method.