Power Converter Control: A Nonlinear Time-Domain Framework to Optimize Efficiency and Performance

University Of Illinois At Urbana-Champaign, Urbana IL

Investigators

Abstract

Switching converters are large-signal, nonlinear systems that emphasize efficient energy processing. This proposal will develop advanced nonlinear large-signal digital control methods to supplant linear small-signal analog controllers in switching power converters. Intellectual Merit - Digital controllers have deeply penetrated certain power electronics markets, such as motor drives, where the incremental cost is small relative to the value provided. The objectives of this project are to develop new performance metrics, address model uncertainty, and explore energy optimization to expand the application of digital nonlinear techniques. The new metrics will relate more directly to nonlinear control theory, rather than linear system theory, and will drive new control methods. Model uncertainty can be substantial in power converters and is complicated by quantization effects when digital controllers are used. Energy optimization is a natural outgrowth of power electronics, where advanced digital techniques can be used to quickly force a solar panel to its maximum power output or a motor to its maximum efficiency point. The resulting set of nonlinear and digital techniques will provide design and analysis tools for a broad range of applications. Broader Impact- The broader impacts of this project are improved energy efficiency and energy processing, the involvement of students at many levels in advanced research, and the opportunity to take switching power conversion in new directions for the future. The methods to be studied have profound implications for energy efficiency, potentially 2.5% of national energy consumption. In addition to graduate research, the project team has a strong history of undergraduate research and large student projects. Recent projects have included about fifty undergraduates in team projects and thirteen undergraduates in independent research. The control approaches open the possibilities for functional optimization of efficiency, performance, and other factors that greatly enhance electrical energy processing.