Real-time system identification for load monitoring and transient handling of Dc-Dc supplies

Grant E. Pitel, Philip T Krein

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Switching mode power supply (SMPS) behavior depends entirely upon known component values and often unknown load impedances. Load change produces voltage over- and undershoots and a design-point shift. In most cases, the controller is blind to these shifts. However, load knowledge is an essential design parameter and should precede optimal and adaptive control techniques. System identification algorithms implemented on digital processors open new opportunities in control and system performance. Typical studies in SMPS identification are based on offline steady-state measurements that form nonparametric frequency-domain models. This paper investigates real-time system identification algorithms that generate parametric models - a practical form for pole-placement and root-locus design. Two key metrics are introduced: parameter error and convergence time to describe algorithm accuracy and speed, respectively, after abrupt parameter changes - a common SMPS load scenario. Hardware and simulation results will show that an algorithm called recursive least squares, in its most basic form, could reasonable approximate the input-to-inductor current plant for static loads during startup, but not for abrupt load steps.

Original languageEnglish (US)
Title of host publicationPESC '08 - 39th IEEE Annual Power Electronics Specialists Conference - Proceedings
Pages3807-3813
Number of pages7
DOIs
StatePublished - Sep 29 2008
EventPESC '08 - 39th IEEE Annual Power Electronics Specialists Conference - Rhodes, Greece
Duration: Jun 15 2008Jun 19 2008

Publication series

NamePESC Record - IEEE Annual Power Electronics Specialists Conference
ISSN (Print)0275-9306

Other

OtherPESC '08 - 39th IEEE Annual Power Electronics Specialists Conference
CountryGreece
CityRhodes
Period6/15/086/19/08

Keywords

  • Dc-dc power conversion
  • Identification
  • Parameter estimation
  • Predictive control

ASJC Scopus subject areas

  • Modeling and Simulation
  • Condensed Matter Physics
  • Energy Engineering and Power Technology
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Real-time system identification for load monitoring and transient handling of Dc-Dc supplies'. Together they form a unique fingerprint.

Cite this