Online model-predictive thermal management of inverter-fed electrical machines

• Echtzeitfähiges modelprädiktives Thermomanagement von umrichtergespeisten elektrischen Maschinen

Qi, Fang; de Doncker, Rik W. (Thesis advisor); Hameyer, Kay (Thesis advisor)

Aachen : ISEA (2018, 2019)
Book, Dissertation / PhD Thesis

In: Aachener Beiträge des ISEA 126
Page(s)/Article-Nr.: 1 Online-Ressource (viii, 154 Seiten) : Illustrationen, Diagramme

Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2018

Abstract

All electrical machines generate losses in the form of heat generation, which leads to a constant change of operating temperature of the machines. This temperature variation jeopardizes machine control quality, and could cause damage to the machine at high temperatures. This thesis introduces a generic thermal modeling methodology based on space-resolved lumped parameter thermal network, which allows an automated thermal modeling procedure for electrical machines with scalable model complexity. The resulting high accuracy real-time temperature observer strongly improves torque accuracy of an induction machine. Based on the accurate thermal model, a model predictive algorithm is proposed which dynamically limits the operating range of electrical machines to achieve maximum thermal utilization. Both results are experimentally validated on test bench.