Power Allocation Strategy Based on Decentralized Convex Optimization in Modular Fuel Cell Systems for Vehicular Applications

Abstract

Recently, modular powertrains have come under attentions in fuel cell vehicles to increase the reliability and efficiency of the system. However, modularity consists of hardware and software, and the existing powertrains only deal with the hardware side. To benefit from the full potential of modularity, the software side, which is related to the design of a suitable decentralized power allocation strategy (PAS), also needs to be taken into consideration. In the present study, a novel decentralized convex optimization (DCO) framework based on auxiliary problem principle (APP) is suggested to solve a multi-objective PAS problem in a modular fuel cell vehicle (MFCV). The suggested decentralized APP (D-APP) is leveraged for accelerating the computational time of solving the complex problem. Moreover, it enhances the durability and the robustness of the modular powertrain system as it can deal with the malfunction of the power sources. Herein, the operational principle of the suggested D-APP for the PAS problem is elaborated. Moreover, a small-scale test bench based on a light-duty electric vehicle is developed and several simulations and experimental validations are performed to verify the advantages of the proposed strategy compared to the existing centralized ones.