Multi-Objective Optimized Designing of Hybrid Powertrain for the Diesel-Electric Locomotive

The key to designing a hybrid locomotive is to determine suitable component sizing. For this reason, this paper proposes a new bi-level optimization design method with two nested loops. The improved multi-objective particle swarm optimization (IMOPSO) is used in the outer-loop to optimize component sizing. The Pontryagin minimum principle (PMP) is adopted in the inner-loop to design the energy management strategy for each component sizing derived from the outer-loop. Meanwhile, to rapidly get the numerical solution of the PMP strategy for different component sizing, this paper proposes an adaptive dichotomy and applies it to the inner-loop. The simulation results show that compared with the general MOPSO, the IMOPSO reduces the generational distance (GD) by 46.4% and increases the maximum spread (MS) by 5%. In addition, the average number of iterations required for the adaptive dichotomy is reduced by 23.3% compared to the traditional dichotomy.