Self-sensing permanent magnet machine

  • Tianhao Wang

Student thesis: PhD Thesis


This thesis looks at the saliency-based self-sensing control of permanent magnet synchronous machines (PMSM) and a novel machine configuration is proposed to improve the self-sensing performances. In recent time, PMSM drives have been steadily gaining popularity and have widespread applications in industry due to its benefits such as high power density, good dynamic performance and high efficiency. Self-sensing drives are superior to conventional drives in applications where the reliability and the cost are of important factors. Machine saliency is utilized for rotor position tracking during the start-up and the low speed operation when Back-EMF components are not detectable. For conventional PMSM machines, however, the saliency of Interior Permanent Magnet (IPM) machine is heavily affected by saturation effects under loaded operation; for the case of Surface Mounted Permanent Magnet (SMPM) machines, the saliency is not apparent and hard to detect. Hence the rotor position signals are relatively small or even undetectable at specific operation points, and these are the main challenges of PMSM drive self-sensing controlled at the low speed. Addition of a novel saliency modulation rotor end (SMRE) structure to the end of a conventional PMSM rotor to improve the self-sensing capability is proposed. The SMRE provides an additional space anisotropic to the rotor. The saliency modulation of the rotor end is electrically asynchronous with the machine`s rotating reference frame. Therefore, the machine saliency provided by SMRE is not affected by saturation effects under loaded operation when high frequency injection scheme is adopted in low speed ranges. In addition, for the medium and high speed range, the rotor position can be tracked without superposed injection as the saliency modulation can be achieved by taking the fundamental voltage as the carrier signal. A genetic algorithms (GA) optimization environment combined with the finite element analysis (FEA) enables to obtain optimized rotor end geometry for the maximum modulation signal and minimum total harmonic distortions (THD). The expected self-sensing performance is validated by a prototype machine and is compared with conventional PMSMs in experimental tests.
Date of Award2 Jul 2017
Original languageEnglish
Awarding Institution
  • Univerisity of Nottingham
SupervisorChris Geradag (Supervisor), John Xu (Supervisor) & He Zhang (Supervisor)


  • Self-sensing
  • permanent magnet synchronous machines

Cite this