System Optimization of Multistack and Multimotor Powertrain for Fuel Cell Electric Vehicles

Abstract

Multistack and multimotor powertrain systems have significant potential for improving the efficiency and performance of fuel cell electric vehicles (FCEVs) compared to conventional powertrain systems. To achieve a superior powertrain system, the major components such as the stack, motor, and transmission of the multistack and multimotor systems should be optimized. To analyze the energy efficiency and dynamic performance of the FCEV, an FCEV analysis model was developed. This model included a two-stack and two-motor powertrain system (2S2M) employing a stack power and motor torque distribution strategy. An optimization problem was formulated with stack transition power, motor torque distribution, and transmission gear ratios as the optimization variables and hydrogen consumption and acceleration time as the objectives for efficiency and performance measures, respectively. An artificial neural network (ANN) model-based optimization method was used to address the computational burden of multiobjective optimization. The optimization results highlighted the Pareto front for the FCEV employing 2S2M, showing a trade-off relationship between the efficiency and performance of the FCEV. Compared to the conventional powertrain system, the 2S2M can reduce hydrogen consumption and acceleration time by up to 7.9% and 6.2%, respectively. An analysis of the distribution of optimal solutions and a comparison of the Pareto fronts for each optimization variable highlighted the necessity for the proposed system optimization method. Furthermore, a comparison between the FCEV and ANN models in terms of computational time for the optimization demonstrated the effectiveness of the ANN model-based multiobjective optimization.