Automated averaging techniques for power electronic-based systems

Eric Allen Walters, Purdue University

Abstract

Average-value models are commonly used in the design and analysis of power electronic based systems. There are numerous averaging methodologies that have been developed that are applicable to specific classes of power electronic circuits including state-space averaging, canonical cell averaging, switch averaging, sampled-data modeling, and multi-rate averaging. Historically, the development of each averaging technique and the extensions thereof was motivated by a specific converter topology for which previous averaging techniques were inaccurate or cumbersome. In general, when a new converter topology is encountered, it is typically not clear as to which averaging technique or extension should be used for that converter. Consequently, substantial analytical effort is generally required to develop and implement the corresponding average-value model. In this research, three automated averaging techniques have been developed that differ in complexity and/or the range of applicability. The first of these techniques is automated state-space averaging that is shown to produce accurate small- and large-displacement averaged models for PWM controlled converters and inverters in which the switching intervals are functions only of control inputs. The second technique extends automated state-space averaging to accurately predict low-frequency small-displacement responses for converters in which the switching intervals are state-dependent. Finally, an observer-based automated averaging method is set forth that accurately portrays both small- and large-displacement responses for both classes of converters. Each averaging technique is verified by comparison with established analytical and numerical methods.

Degree

Ph.D.

Advisors

Wasynczuk, Purdue University.

Subject Area

Electrical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS