Maximum Efficiency Control of a Step Skewed Ferrite-Assisted Synchronous Reluctance Machine Accomplished by Quadratic Interpolation
Authors
Pedersen, Kristian Buchwald ; Severinsen, Frederik Brath
Term
4. term
Education
Publication year
2022
Submitted on
2022-05-30
Abstract
Denne afhandling, udført i samarbejde med Grundfos, udvikler og afprøver en styringsmetode til at maksimere virkningsgraden af en trin-skrå ferrit-assisteret synkron reluktansmaskine (FASynRM) til pumpeanvendelser, vurderet ud fra Energy-Efficiency-Index (EEI). Udgangspunktet er, at konventionel MTPA-styring primært minimerer kobbertabet, mens maksimal virkningsgrad (ME) kræver, at både kobber- og kernetab adresseres. Maskinens parametre (statorfasemodstand, PM-fluks og d-/q-akse-induktanser) blev målt, og en FOC-struktur blev implementeret, hvor strømmens vinkel bestemmes af en Maximum Efficiency Algorithm (MEA) baseret på kvadratisk interpolation. Der diskuteres praktiske udfordringer som demagnetiseringsgrænser for ferritmagneterne og lokalisering af rotorens d-akse, hvor BEMF-måling bruges til at korrigere for stabile parkeringspositioner forskudt fra d-aksen. I eksperimenterne blev MEA undersøgt via tre tilgange: (1) aktiv effekt-beregning med inverter-spændingsfejlkompensation (fandt ikke inputeffekten korrekt), (2) inputeffekt-LUT’er afledt af manuelle målinger (validerede MEA), og (3) en tabsestimeringsprofil, der kombinerer kobbertab med lineært estimerede kernetab som funktion af q-aksestrømmen ved 2000, 2500 og 3000 rpm; alle kunne finde ME-vinklen, når startintervallet var valgt korrekt. Samlet viste EEI og årligt energiforbrug kun ubetydelige forskelle mellem ME-, MTPA- og konstant vinkel-styring, så det anbefales at vælge metode efter implementeringsindsats og antal driftsbetingelser frem for forventet EEI-gevinst.
This thesis, conducted in collaboration with Grundfos, develops and evaluates a control approach to maximize the efficiency of a step-skewed Ferrite-Assisted Synchronous Reluctance Machine (FASynRM) for pump applications, assessed using the Energy-Efficiency Index (EEI). While conventional MTPA mainly minimizes copper losses, maximum efficiency (ME) requires addressing both copper and core losses. The machine’s parameters (stator phase resistance, PM flux linkage, and d/q-axis inductances) were measured, and a Field-Oriented Control scheme was implemented in which the current angle is set by a Maximum Efficiency Algorithm (MEA) based on quadratic interpolation. Practical challenges are treated, including demagnetization limits for ferrite magnets and rotor d-axis localization, where BEMF measurements correct for stable parking positions offset from the d-axis. Experiments investigated MEA through three routes: (1) active power calculation with inverter voltage error compensation (did not yield correct input power), (2) input power look-up tables derived from manual measurements (used to validate MEA), and (3) a loss estimation profile combining copper losses with linearly estimated core losses versus q-axis current at 2000, 2500, and 3000 rpm; all could locate the ME angle when the initial angle interval was chosen properly. Overall, EEI and annual energy consumption differences between ME, MTPA, and constant-angle control were negligible, so the preferred method should be selected based on implementation effort and the number of operating conditions rather than expected EEI gains.
[This summary has been generated with the help of AI directly from the project (PDF)]
Documents