Development of A High-Voltage (1200V) Powertrain System for eVTOL Aircraft Using Wide-Band-Gap Semiconductors

With the rapid advancement of electric vertical take-off and landing (eVTOL) aircraft technology, there is a growing demand for more efficient and powerful eVTOL capable of performing complex tasks with extended flight durations. Chinese eVTOL aircraft systems, such as XPENG AEROHT and E-Hang, operate at 400V electrical structures using low-voltage silicon-based semiconductors. International eVTOL companies move forward earlier and adopt a higher voltage powertrain to 800V, such as Joby, Archer, and Beta Technologies. The higher voltage powertrain will offer significant benefits in terms of charging rates, efficiency, and weight reduction. At present, the industry has not been able to break through to more than 1000V due to technical limitations.

To address these challenges, the project proposes a pioneering 1200V architecture for eVTOL powertrain systems, encompassing fast chargers,battery management systems, converters, and machines. Higher voltage systems using next-generation wide-band-gap semiconductors and new power electronics technology will demonstrate advantages in charging rates, power density, and efficiency, which will further enhance the capabilities for heavy and long-distance eVTOL applications.

In this project, we mainly conduct research for the intelligent motor control for eVTOL aircraft. Inspired by the significant speed-drop curves observed in UAVs’ propulsion motors under strong wind disturbances, our team has pioneered the integration of machine learning into motor control. We proposed an “Anti-Disturbance Zero-Speed-Drop Intelligent Control Algorithm” and validated its effectiveness on the propeller test bench. The algorithm achieves zero-speed-drop performance even under sudden 11-level storm disturbances (30 m/s wind speed), overcoming the long-standing technical bottleneck of existing motor control methods in extreme disturbance scenarios. This breakthrough addresses two critical challenges in applying machine learning methods to motor control: interpretability and real-time performance, exemplifying the cross-disciplinary synergy between artificial intelligence and conventional engineering.

Building on this foundation, our team further developed an adaptive zero-speed-drop algorithm library for multi-level wind disturbances and engineered a dedicated electronic speed controller. This technology holds significant potential for motor control in low-altitude aircraft, a cornerstone of the emerging “low-altitude economy.”