Open Collections

Large-signal transient control in power electronics : an average-geometric framework

University of British Columbia

Switch-mode power converters are a fundamental component of modern power systems; they are ubiquitous in renewable energy applications, electric vehicles, battery chargers, and power supplies. Controllers are an essential component in power converters as they improve the converter's dynamic behaviour during transients and in steady-state. For decades, the power electronics industry has preferred to utilize controllers based on converter small-signal analysis due to their low implementation requirements and in spite of their dynamic performance limitations and global stability issues. On the other hand, excellent dynamic response and global stability are achieved by non-linear boundary controllers based on state-plane analysis, which usually have much higher implementation requirements. This thesis focuses on the dynamic performance improvement of power converters by incorporating state-plane concepts while maintaining low implementation requirements to facilitate the large-scale adoption of the technology. By combining traditional averaging modelling tools with state-plane analysis, the unified Average Natural Trajectories (ANTs) are obtained to accurately model the large-signal dynamic behaviour of the fundamental topologies (buck, boost, and buck-boost). As a result, the proposed framework establishes the foundation for several dynamic performance improvement efforts introduced in this thesis. Employing the ANTs as a large-signal model, the theoretical limits of dynamic performance are defined and used to develop a powerful benchmarking tool, providing great value for design engineers. Furthermore, a unified controller based on the ANTs model is developed for the fundamental topologies in this work. This controller features a predictable large-signal response and outstanding dynamic performance while maintaining low implementation requirements. The ANTs modelling approach is also extended to photovoltaic applications to develop an extremely fast maximum power point tracking method for scenarios that include rapidly changing environmental conditions. Finally, the concept of dual-loop geometric control is introduced by combining state-plane analysis with an industry-standard dual-loop control structure, thereby bridging the gap between industrial applications and state-plane controllers. The concepts introduced in this thesis are supported by thorough mathematical analysis and validated by extensive simulation and experimental results. This thesis significantly contributes to the advancement of the field of modelling and control for power electronics.

Text

2020-04-01

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/73884

Electrical and Computer Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/