Dynamic Modeling and Solving Methods for Multi-Train Energy-Efficient Operation and Network Voltage Stability

Abstract

Freight trains operate in dynamic environments and exhibit time-varying behavior, making static mechanistic models inadequate for capturing these changes. This often results in impractical predictions of train operational states and optimization outcomes. To facilitate planning in such operational conditions, this paper proposes a dynamic modeling method to assess energy consumption and the voltage of traction power supply system (TPSS), and a large-scale adaptive multi-strategy multi-objective competitive swarm optimization algorithm (LA-MOCSO) for solving dynamic optimization challenges. Specific, a mechanistic “train-track-power grid” (TTP) model is first built to calculate power flow and TPSS voltage during multiple train operations. Second, a hybrid modeling approach that combines the mechanistic model and data-driven models is proposed to account for variations in train and environmental characteristics, and a multi-objective optimization model is established aimed at improving energy-efficiency and voltage stability of TPSS. Then, to tackle the complexities of the multi-objective optimization problem, an LA-MOCSO algorithm is proposed, which can be applied to solve the large-scale optimization problem of multi-train long-distance routes. Finally, the high accuracy of the dynamic model was validated with measurement data; the performance and computational efficiency of LA-MOCSO was verified through five algorithms; the comprehensive optimization method can, through the allocation and utilization of regenerative braking energy, further reduce substation energy consumption and maintain grid voltage stability.