Design Trade-off Analysis and Optimisation of a High-density Active Shunt Regulator for Aircraft

Student thesis: Doctoral ThesisDoctor of Philosophy (PhD)


Moving towards More Electric Aircraft (MEA) demands for high efficiency and high power density power electronics. There are two important areas that are pushing the boundaries of future power converters: advanced converter topologies (e.g. three-level topology) and emerging power semiconductor devices (e.g. Wide Band Gap power devices). The benefits of these new technologies are still subject to quantified analysis from the system
level point of view over a particular application.
The aim of this work is to investigate the design trade-offs and optimisation of a 27 kVA three-phase Active Shunt Regulator (ASR) system regarding three main design variables: topology, power device and switching frequency. The emerging technologies, the three-level T-type topology and Silicon Carbide (SiC) devices, are evaluated against the state-of-the-art commercial solution, i.e. Silicon IGBT and two-level topology. To realize a holistic view of the system, not only the active power devices, but also the passive components require designing
and accurate modelling. Passive components considered in this work include the heatsinks, the filter inductors and the DC-link capacitors, which account for a substantial contribution to the power loss and volume/weight of the converter system. Power loss models, design procedures and volume/weight models are established for individual components of the ASR system. As the most challenging part, the core loss of inductors is thoroughly investigated and evaluated. Additionally, the neutral point voltage balancing issue in three-level converters is addressed in this work when the ASR system operates at a low power factor.
The developed mathematical optimisation tool shows a prediction on the achievable performance of the converter and the quantified design trade-offs. For example, increasing the switching frequency reduces the size of filter components while it requires larger heatsink for increased switching loss. Prototypes are built to validate the theoretical models and realize the optimal design identified.
Date of Award25 Jun 2019
Original languageEnglish
Awarding Institution
  • The University of Bristol
SupervisorXibo Yuan (Supervisor) & David A W Barton (Supervisor)

Cite this