Date of Award

12-2025

Degree Name

Master of Science in Engineering

First Advisor

Sandun S. Kuruppu, Ph.D.

Second Advisor

Damon A. Miller, Ph.D.

Third Advisor

Richard T. Meyer, Ph.D.

Keywords

Extremum-seeking control, permanent magnet motors, real-time systems

Access Setting

Masters Thesis-Open Access

Abstract

The inherent advantages of permanent magnet synchronous machines (PMSMs), such as high efficiency, has led to their frequent use in industry, including safety-critical applications, such as transportation electrification. The accurate measurement of the rotor position is key to achieving efficient and reliable field-orientation-based torque regulation. With the high reliance on the rotor position measurement, position sensor fault diagnosis and mitigation facilitate a layer of safety. Conventional methods for compensating position sensor offset errors in electric drives typically depend on accurate knowledge of motor parameters. However, these parameters are often subject to variation due to temperature changes, system aging, and other operational uncertainties, which can compromise the effectiveness of fault detection and compensation strategies. This thesis proposes a model-free extremum-seeking control approach for self-healing compensation of position sensor offset errors. The method eliminates the need for motor parameter dependence by leveraging online performance optimization, enabling robust error detection and correction under varying system conditions. The theoretical framework for the proposed approach is presented along with simulation and experimental validation. Furthermore, this study proposes a method to quantify the Position Sensor Offset Error in extremum-seeking-driven non-salient PMSMs as it is essential for fault severity assessment.

Download

Share

COinS