On the Exactness of an Energy-efficient Train Control model based on Convex Optimization

Abstract

In this paper, we demonstrate the exactness proof for the energy-efficient train control (EETC) model based on convex optimization. The proof of exactness shows that the convex optimization model will share the same optimization results with the initial model on which the convex relaxations are conducted. We first show how the relaxation on the initial non-convex model is conducted and provide analysis to show that the relaxations are convex constraints and the relaxed model is thus a convex model. Subsequently, we prove that the relaxed convex model will always achieve its optimal solution on the initial equality constraints and the optimal solution achieved by convex optimization will be the same as the one obtained by the initial non-convex model and the relaxations applied are exact. A numerical verification has been conducted based on a typical urban rail system with a steep gradient. The results of this paper shed lights on further applications of convex optimization on energy-efficient train control and relevant areas related to operation and control of low-carbon transportation systems.