Noise-Adaptive Multimode Online Energy Management for PEMFC/Battery Hybrid UAVs

Hydrogen/battery hybrid unmanned aerial vehicle (UAV) flights present unique challenges to the adaptability of energy management strategy (EMS) due to dynamic operating conditions (altitude, temperature, and humidity) and diverse flight modes (takeoff, cruising, and maneuvering). In this article, a novel multimode EMS is proposed. First, inspired by the variational Bayesian (VB) approach, a noise-adaptive parameter identification method is introduced to monitor the fuel cell (FC) characteristics in-flight. The identification results provide an online reference for energy management. Subsequently, a case recognition logic categorizes the flight mode into cruising and noncruising based on flight power variation. An online rule-based method is deployed for the noncruising case to prioritize system response, and a novel equivalent consumption minimization strategy (ECMS) is used to maximize system endurance during cruising. Extensive ground tests are conducted with a FC in a constant temperature and humidity chamber, and a flight test is carried out on a self-developed 3 kW FC/battery hybrid UAV. Experimental results show that the proposed method outperforms classic EMSs in terms of system efficiency and reduced system stress.