Fuel cell hybrid electric vehicles: experimental characterization, design and control for improving energy efficiency and performance

Nowadays climate change represents a pressing challenge, with the transportation sector standing as one of the major contributors to global greenhouse gas emissions. Addressing this issue requires the development of sustainable mobility solutions, among which Fuel Cell Hybrid Electric Vehicles offer a promising alternative owing to their ability to combine high energy efficiency, extended range, and zero tailpipe emissions. Indeed, hydrogen technologies are effective means towards the road of energy rationalization, since it enables renewables flexibility and grid stabilization. However, a key challenge is raised by the design of such hybrid electric powertrains, generally comprising a fuel cell stack and a battery pack, since proper sizing and control are fundamental in unlocking the full potential of these technologies and a co-optimization approach is required due to their strong interdependence. The goal of the present work is the development of a structured methodology for the powertrain design process from a technological standpoint. The methodology employed involves an experimental and modeling miscellaneous approach, applied on a non-conventional hybrid electric light quadricycle due to the growing interest in micromobility, supporting decarbonization in high-populated urban areas. Experimental characterization of real-scale powertrain components, i.e. FCS, battery pack, and DC/DC converter, has been conducted in the laboratory, whose data have then been used to develop and validate a detailed simulation platform within MATLAB/Simulink/Simscape framework, capable of accurately replicating vehicle behavior under a variety of operating conditions. This simulation platform was further integrated into a Hardware-inthe-Loop system, enabling real-time analysis of the real powertrain components for accurate performance evaluation. Additionally, a scale-up approach has been implemented into the platforms to assess the impact of different FCS sizes on vehicle performance. In order to ensure efficient behavior of the powertrain, a novel EMS based on Driving Pattern Recognition technique has been proposed, able to dynamically adapt to the actual driving conditions and operate the powertrain in the optimal efficiency regions, owing to prior training through Dynamic Programming optimal results. The case study has involved four FCS size scales (scale-up factors of 1, 1.5, 2 and 2.5), which have been tested on both HIL and simulation platforms, thus finally determining the proper choice as design process outcome. The simulation results have demonstrated high accuracy in estimating the real behavior of the components obtained by HIL test bench, confirming the improvement of the vehicle performance as the FCS size increases, thus proposing the highest-powered system as the most suitable for the design choice. Indeed, the FCS average net efficiency increased from 46.1 % for the baseline to 58.6 % for 𝑓𝑆𝑐𝑎𝑙𝑒−𝑈𝑝 = 2.5 and the estimated vehicle range has exhibited an improvement of nearly 19 % (from 81.1 km to 96.7 km). Moreover, the obtained results demonstrate the fundamental need of addressing the design process in a co-optimized approach, achieving increased performance of nearly 8-9 % when adopting the novel EMS with respect to the baseline Fuzzy Logic controller.