Robust stator flux and load torque estimations for induction motor drives with EKF-based observer

DEMİR R.

ELECTRICAL ENGINEERING, vol.105, no.1, pp.551-562, 2023 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 105 Issue: 1
  • Publication Date: 2023
  • Doi Number: 10.1007/s00202-022-01717-y
  • Journal Name: ELECTRICAL ENGINEERING
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Compendex, INSPEC, DIALNET
  • Page Numbers: pp.551-562
  • Keywords: Induction motors drive, Extended Kalman filter, DTC, Robust estimation, EXTENDED KALMAN FILTER, SPEED-SENSORLESS CONTROL, REFERENCE ADAPTIVE SYSTEM, RESISTANCE ESTIMATION, PARAMETER-ESTIMATION, ROTOR RESISTANCE, PERFORMANCE, CONSTANT
  • Kayseri University Affiliated: Yes

Abstract

This paper introduces an extended Kalman filter (EKF)-based new robust state and parameter estimation for induction motor (IM) drives. The EKF observer uses a proposed reduced IM model to estimate the rotor angular speed, stator fluxes, and the load torque including viscous friction term independently from the rotor resistance, rotor, stator, and magnetizing inductances. The EKF observer with the reduced IM model uses the rotor angular speed in the measurement equations, whereas the other existing methods use the stator currents or both the rotor speed and the stator currents. Thus, the estimations of the state and parameters in the EKF observer are carried out according to speed error. The stator flux-based IM model used in the EKF observer designed in this paper does not include the stator currents as the state. Thus, due to the use of the proposed reduced IM model in the EKF observer, the design stages of the observer and the computation time are reduced. The average execution time of each iteration for the EKF observer which uses the proposed reduced IM model has been measured as 2.1 mu s in real-time experiments. The estimation performance of the proposed EKF observer has been first implemented and tested in simulations and then tested in real-time experiments. The experimental results are obtained by the dSPACE DS1104 controller board, connector panel, and ControlDesk software. When the simulation and experimental results obtained from a wide speed range are considered, the validity of the EKF observer which uses the reduced IM model is quite satisfactory.