A multi-objective optimal sizing scheme for hybrid traction power supply systems onboard shunting locomotive

Abstract

Hybrid traction power supply systems (HTPSSs) are increasingly being utilized to power electrical vehicles for passenger or for hevay duty, electrified rolling stocks, more-electric ships, etc. Sizing of components is critical for the HTPSS’s dynamic, economical, and environmental performance. However, many HTPSSs have been sized based on engineering experience or simple calculation, which results in wastage of resources and environmental pollution. This paper proposes a multi-objective optimal sizing scheme for HTPSSs for shunting locomotives, aimed at minimizing costs over their life cycles, achieving lightweight goals, and reducing fuel concumption (FC). A multi-objective cost function is consequently constructed, including the above three factors. To identify the best optimization algorithm, the multi-objective cost function with pre-defined energy management strategies (EMS) is initially optimized using three different heuristic algorithms: genetic algorithm (GA), simulated annealing (SA), and particle swarm optimization (PSO) algorithm. The results indicates that the PSO algorithm is the most efficient method for solving this multi-objective optimization problem. To further reduce the FC, a PSO-based two-layer optimization model is presented, which is able to optimize the sizing and the key thereshold of EMS simultaneously, thus, addressing the strong coupling between the two. The result indicates that the system's performance is significantly enhanced compared to the initial sizing: economic costs are reduced by 2,708,800 CNY (8.33%), the total weight of the system is decreased by 1267.39 kg (17.39%), and the FC is reduced by 304.77 t (12.1%) over the 32-year life cycle of the shunting locomotive.

Keywords

• Multi-objective cost function
• Optimal sizing
• PSO
• Shunting locomotive