Speed Sensorless State Estimation for Induction Motors: A Moving Horizon Approach

    •  Zhou, L.; Wang, Y., "Speed Sensorless State Estimation for Induction Motors: A Moving Horizon Approach", American Control Conference (ACC), DOI: 10.1109/ACC.2016.7525249, July 2016, pp. 2229-2234.
      BibTeX Download PDF
      • @inproceedings{Zhou2016jul2,
      • author = {Zhou, L. and Wang, Y.},
      • title = {Speed Sensorless State Estimation for Induction Motors: A Moving Horizon Approach},
      • booktitle = {American Control Conference (ACC)},
      • year = 2016,
      • pages = {2229--2234},
      • month = jul,
      • doi = {10.1109/ACC.2016.7525249},
      • url = {}
      • }
  • MERL Contact:
  • Research Area:


This paper investigates the speed sensorless state estimation problem for induction motors. Aiming at developing new state estimation means to improve the estimation bandwidth, this paper proposes various moving horizon estimation (MHE)- based state estimators. Applying MHE for induction motors is not straightforward due to the fast convergence requirement, external torque disturbances, parametric model errors, etc. To improve speed estimation transient performance, we propose an MHE based on the full induction motor model and an assumed load torque dynamics. We further formulate an adaptive MHE to jointly estimate parameters and states and thus improve robustness of the MHE with respect to parametric uncertainties. A dual-stage adaptive MHE, which performs parameter and state estimation in two steps, is proposed to reduce computational complexity. Under certain circumstances, the dual-stage adaptive MHE is equivalent to the case with a recursive least square algorithm for parameter estimation and a conventional MHE for state estimation. Implementation issues and tuning of the estimators are discussed. Numerical simulations demonstrate that the proposed MHE estimators can effectively estimate the induction motor states at a fast convergence rate, and the dualstage adaptive MHE can provide converging state and parameter estimation despite the initial model parametric errors.