A Lean Approach to Zero and Low-Speed Sensorless Control of PMaSynRMs


Creative Commons License

Tap A., Akgül K., Ergenç A. F., Yılmaz M., Ergene L.

IEEE Access, vol.11, pp.135406-135422, 2023 (SCI-Expanded) identifier

  • Publication Type: Article / Article
  • Volume: 11
  • Publication Date: 2023
  • Doi Number: 10.1109/access.2023.3337524
  • Journal Name: IEEE Access
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, INSPEC, Directory of Open Access Journals
  • Page Numbers: pp.135406-135422
  • Keywords: extended EMF, High-frequency voltage injection, permanent magnet assisted synchronous reluctance motor, sensorless control
  • Istanbul Technical University Affiliated: Yes

Abstract

Sensorless control of permanent magnet motors in full speed range is indispensable for home appliance applications due to cost, robustness and maintenance requirements. Zero and very low-speed sensorless control is a challenging area for model-based position observers. The common methodology for position detection at zero and low speeds is high-frequency signal injection methods due to their independence from model parameters. The major drawback of such methods is the signal processing burden for demodulation of resultant high-frequency currents. In this study, a simplified filtering scheme with a reduced number of filters is proposed for high-performance zero and low speed control of the Permanent Magnet Assisted Synchronous Motors for washing machine applications with a low-cost microcontroller. The proposed method employs the rotating voltage vector injection using a single band-pass filter combined with N-sample averaging. This approach reduces system complexity by reducing the number of filtering stages to just one while providing robustness to estimations under fast current and speed gradients. The proposed method leaves 47% CPU overhead (without any optimization) including notch filtering of the dq-axes current feedbacks for better current control and parallel operation with an SMO based Back-EMF observer for switchover at higher speeds. Experimental results show that the proposed method is suitable for a low-cost microcontroller implementation in a washing machine with maximum uncompensated and compensated errors of 2.28 and 0.708 degree electrical respectively. The position estimation algorithm is shown to be suitable for unfavorable operating conditions such as zero speed acceleration under full load, step speed commands, step direction reversals, intermittent loads such as changing drum inertia and pulsating torque at very low speeds. The proposed system is able to track position with position errors up to 0.708 degree electrical and maintain speed control at very low speeds under 30% pulsating load torque.