Permanent magnet synchronous machines terminal attributes for self-sensing and automatic identification

The principal research objective of this thesis is to develop a systematic and easy-to-automate methodology that will enable implementation of carrier-based self-sensing (sometimes referred to as "mechanical position sensorless") control on existing permanent magnet synchronous machines (PMSMs). The need to produce PM synchronous machine drives which are as universal and as easy to use as their induction machine predecessors is currently a major concern of drive manufacturers. Rapid and robust implementation of self-sensing has the potential to replace the classical position sensor used for such drives, which is a key factor in their acceptance. Automatic identification, characterization and mitigation of impediments to PMSM self-sensing are key elements of this methodology.;The first step of this research includes analytical and experimental characterization of the terminal attributes that affect and limit self-sensing in PMSMs. Once identified, mitigation techniques are developed that target each of the specified self-sensing limitations and their integration into the universal PMSM self-sensing position estimator/observer.;In order to provide a comprehensive perspective, pulsating vector-based self sensing is analyzed and compared with rotating vector carrier-based self-sensing for its sensitivity to the specified terminal attributes and to examine the suitability of combining the two techniques for the purpose of automatic identification. Automatic identification of PMSM attributes for self-sensing is a key enabler for integration of mitigation techniques, and must be done without the use of additional sensors or special testing equipment.;A three-step self-sensing implementation methodology based on conventional and newly proposed methods is the capstone result of this work.