Robust Optimization of Energy-Saving Train Trajectories Under Passenger Load Uncertainty Based on p-NSGA-II
Xing Chen, Kang Li, Li Zhang, Zhongbei Tian
Abstract
Railway electrification has attracted substantial interest in recent years as a key part of the global effort to achieve transport decarbonization. To improve the energy efficiency of train operations, of particular interest is the optimization of train speed trajectories. However, most studies formulate the problem as a single-objective optimization model and do not take into account train mass uncertainty associated with the passenger load variations. This article formulates a biobjective robust optimization model to minimize both the energy consumption and journey time, in which the robustness against the uncertain train mass is considered and viewed as a decision-maker preference. A novel multiobjective optimization algorithm, namely, p-nondominated sorting genetic algorithm-II (NSGA-II), is proposed, incorporating the original NSGA-II and a proposed preference dominance criterion to handle the DM preference. With the proposed p-NSGA-II, not only all solutions will converge to the optimal Pareto front but also solutions with better robustness in the Pareto front will be automatically selected and retained; meanwhile, the spread of the selected solutions is maintained. The effectiveness of the p-NSGA-II to generate a set of performance-robust driving schemes is verified by numerical case studies. The results show that the p-NSGA-II can achieve up to 40.59% robustness improvement compared to the original NSGA-II.