Control and Parameter Identification of a Permanent Magnet Synchronous Motor with a LC-filter
Author
Pedersen, Anders Roland
Term
4. term
Education
Publication year
2013
Submitted on
2013-06-04
Pages
79
Abstract
Denne afhandling undersøger, hvordan en overflademonteret permanentmagnet-synkronmotor (PMSM) kan styres, når den drives af en spændingskildeomformer (VSI) gennem et LC-filter. VSI giver præcis hastighedskontrol, men de høje dv/dt kan medføre lejestrømme, overspændinger ved lange motorkabler og akustisk støj. Et LC-filter dæmper disse effekter, men ændrer samtidig amplituder og faser af spændinger og strømme ved motoren, hvilket komplicerer styringen og kræver kendskab til filterets parametre. Arbejdet udvikler simuleringsmodeller af PMSM, LC-filter og inverter og foreslår en feltorienteret, FOC-lignende styringsstruktur, der regulerer inverterstrømme, motorstrømme, motorspændinger og motorhastighed via frakobling og internal model control-principper. For at understøtte idriftsættelse undersøges, om Recursive Least Squares (RLS) kan identificere LC-filterets parametre ud fra passende testsignaler og testkonfigurationer. I de gennemførte forsøg kunne kun dele af styringen verificeres på grund af tidsbegrænsninger og fejlbehæftede målinger, og den anvendte RLS-tilgang viste sig utilstrækkelig, hvilket indikerer behov for forbedret estimering og yderligere eksperimentel validering.
This thesis examines how to control a surface-mounted permanent magnet synchronous motor (PMSM) when it is driven by a voltage source inverter (VSI) through an LC filter. VSIs enable precise speed regulation, but high dv/dt can cause bearing currents, overvoltages on long motor leads, and acoustic noise. An LC filter mitigates these effects but also alters the amplitudes and phases of voltages and currents at the motor, complicating control and requiring knowledge of the filter parameters. The work develops simulation models of the PMSM, LC filter, and inverter, and proposes a field-oriented, FOC-like control structure that regulates inverter currents, motor currents, motor terminal voltages, and motor speed using decoupling and internal model control principles. To support commissioning, the study investigates whether Recursive Least Squares (RLS) can identify the LC filter parameters from suitable test signals and test configurations. In the available tests, only part of the controller could be verified due to time constraints and measurement issues, and the RLS approach applied proved inadequate, indicating a need for improved estimation and further experimental validation.
[This summary has been generated with the help of AI directly from the project (PDF)]
Documents