Title:
Multi-objective design, optimization, and condition monitoring of high-performance electric machines for electric propulsion
Multi-objective design, optimization, and condition monitoring of high-performance electric machines for electric propulsion
Author(s)
Zhang, Shen
Advisor(s)
Habetler, Thomas G.
Editor(s)
Collections
Supplementary to
Permanent Link
Abstract
The objective of the proposed research is to develop methods for the multi-objective design, optimization, and condition monitoring of electric machines, so as to generate the optimal designs and improve machine robustness for electric propulsion. In particular, the selected high-performance electric machines are the switched reluctance machine (SRM) with a simple and robust structure, and the interior permanent magnet (IPM) machine with a high torque density and efficiency. For SRMs, an active current profiling technique integrated multi-objective analytical design and optimization is proposed to generate the optimal design in terms of multiple performance indices, which is proven to be accurate and time-saving, especially for a large search space with multiple prime design variables. The proposed scheme offers machine designers accurate, handy and convenient initial designs, which can be further verified or fine-tuned if necessary. The optimization process is further developed with advanced machine learning algorithms to accelerate the search process and facilitate the final decision-making process with the self-organizing map and t-SNE algorithm. To monitor the demagnetization property of the closed-loop direct torque controlled (DTC) IPMSM, two nonintrusive high-frequency signal injection schemes are proposed for PM temperature estimation via analyzing the PM electrical high-frequency resistance, which is a byproduct of the eddy current loss induced by the applied high-frequency magnetic field. The developed methods bring practical ways to excite a proper amount of high-frequency current into the stator winding, which leads to a simple, accurate, and nonintrusive permanent magnet thermal monitoring scheme for DTC-controlled IPM machines. The demagnetization properties of the IPM machine under the most commonly observed stator inter-turn short circuit fault is also systematically investigated via simulations and experiments, thereby offering machine designers handy information in evaluating the demagnetization fault-tolerant capability of various IPM machine design candidates.
Sponsor
Date Issued
2019-05-23
Extent
Resource Type
Text
Resource Subtype
Dissertation