Analysis of Advanced Powertrain Architectures for Electric Racing Motorcycles Considering Hybrid Energy Storage and Wide-Bandgap Semiconductors

Abstract

The rapid progress in the fields of energy storage and power electronics offers various promising improvements of conventional powertrain architectures. This thesis investigates the improvement potential of Hybrid Energy Storage Systems (HESSs) and Gallium Nitride (GaN) semiconductors in the powertrain of an electric racing motorcycle, designed for the MotoStudent competition. To ascertain the state of the art, the powertrain architectures of several commercial electric motorcycles and racing prototypes are compared. Subsequently, the powertrain of the current prototype, developed at the University of Oviedo, is explained in detail and the performance is analytically assessed. As first potential improvement, HESSs made of Lithium-Ion (Li-ion) batteries and Ultracapacitors (UCs) are evaluated. An optimization algorithm for the design of the HESS, which considers two semi-active topologies, is introduced. Further analysis juxtaposes the optimization results with pure Li-ion battery storage in terms of weight and volume. In the case of the MotoStudent load profi le, results in favor of pure battery storage are presented. Thereupon, the feasibility of a 500 ARMS GaN traction inverter is analyzed. For this purpose, an electrical and thermal co-simulation is performed in PLECS. In order to accurately model the semiconductor losses, 3D lookup tables are utilized. Based on the results, a minimum number of sixteen parallel devices is determined to meet the requirements. As such high numbers of parallel devices have not been studied in literature yet, a double-pulse test Printed Circuit Board (PCB) is designed to evaluate the feasibility. The measurements carried out demonstrate the full functionality of the design and successful parallelization of sixteen GaN transistors.