Train operation optimization with adaptive differential evolution algorithm based on decomposition

Abstract

The train operation is a multi‐objective optimization process including several operating indexes, such as safety, punctuality, precision stop, comfort, and low‐energy consumption. The speed profile and timetable for trains should be adjusted correspondingly to satisfy the passengers' demands while minimizing energy consumption. To solve this problem, we propose a multi‐objective optimization approach for train operation. First, we develop a single‐particle model of the train and a multi‐objective optimization model for the train operation, which is subject to the constraints such as safety requirement, passenger comfort, and low energy consumption. Second, to obtain the Pareto frontier of train operation, a uniform design multi‐objective adaptive differential evolution algorithm based on decomposition (UMADE/D) is studied and applied to solve the multi‐objective optimization model for the train operation. Finally, we present numerical examples based on the real‐life operation data from the Nanjing Metro Line 1 in Nanjing, China. Three searching methods, namely, the proposed UMADE/D, Non‐Dominated Sorting Genetic Algorithm, and Multi‐Objective Evolutionary Algorithm based on Decomposition, are implemented to find the Pareto frontier of train operation quickly and efficiently. The simulation results show the efficiency of the proposed UMADE/D algorithm, which can provide corresponding operation strategies and satisfy the multi‐objective demands when the train is in different operation state. © 2019 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.