H∞ observer for sensorless velocity control of permanent magnet synchronous motors

Hsin Hung Chou, Stone Cheng, Chia Min Ting

Research output: Contribution to conferencePaperpeer-review

Abstract

This paper mainly adopts robust control to realize a speed controller for sensorless permanent magnet synchronous motor (PMSM). To control speed, the robust control estimator is used to estimate the motor back-electromotive force (back-EMF) and then estimate the magnetic pole angle θ e corresponding to motor rotation, as so to send a feedback to the current loop for vector control. This method further converts the magnetic pole angle into the rotor speed for speed control. In the system of the motor control loop, due to motor parameter uncertainty and external load interference, the motor angle estimation error cannot be guaranteed to be converged within a tolerable range, thereby causing system instability. In order to improve estimator robustness, the H∞ observer method is derived to realize the speed controller for sensorless PMSM. The H∞ observer relies on the estimated current and actual current values to estimate the magnetic pole angle of the motor. The H∞ observer guarantees that the gain value for the worst-case estimation error is within the index range of design performance. This paper proves the convergence of the H∞ observer and utilizes matlab simulations to verify results. The simulation results show that the H∞ observer-based speed controller for senseless PMSM yields good performance.

Original languageEnglish
StatePublished - 16 Aug 2010
Event10th International Conference on Motion and Vibration Control, MOVIC 2010 - Tokyo, Japan
Duration: 17 Aug 201020 Aug 2010

Conference

Conference10th International Conference on Motion and Vibration Control, MOVIC 2010
Country/TerritoryJapan
CityTokyo
Period17/08/1020/08/10

Keywords

  • H∞ observer
  • PMSM
  • Robust control
  • Sensorless

Fingerprint

Dive into the research topics of 'H∞ observer for sensorless velocity control of permanent magnet synchronous motors'. Together they form a unique fingerprint.

Cite this