Iterative Tuning Feedforward Speed Estimator for Sensorless Induction Motors

For speed sensorless induction motors under field-oriented control (FOC), where the motor speed and angle are not measured, the speed control tracking bandwidth is mainly limited by the convergence rate of the state estimator. Prevailing speed-sensorless induction motors suffer significant performance degradation from removing the encoder, which limits their applications to fields requiring low or medium performance. This paper studies a new estimation approach for induction motors, aiming at improving improving the estimation bandwidth of sensorless induction motors, and thus enabling them for higher bandwidth drives. In the speed-sensorless estimation for induction motors, the classic model reference adaptive system (MRAS) approach [1, 2, 3, 4] was initially studied and remains appealing until today. Although the design was simple, this method often suffers from slow converging due to the adaptive estimation. Through the years, numerous estimation methods have been studied for induction motors, such as sliding mode observer [5, 6], extended Kalman filter (EKF) methods [7, 8], moving horizon estimation (MHE) methods [9], etc. In these designs the rotor’s mechanical equition is often not included in the estimator model, based on the assumption that the mechanical dynamics is slow compared to the electrical dynamics. This assumption can significantly simplify the estimator design and allow the estimation to proceed without knowing the motor’s mechanical parameters, but often results in slow transient. In this work, we propose a new induction motor state estimation method with the rotor’s mechanical dynamics included, targeting at improving the speed estimation convergence rate and thus improve the speed tracking bandwidth of the sensorless induction motor. In the proposed estimator design, the estimations of rotor flux and the rotor speed are separated into two sequential steps, where the flux estimation is achieved by a linear filter, and the speed estimation uses a combination of feedforward and feedback. In order to address the difficulty of unknown rotor inertia and load torque, an iterative tuning method was used to automatically tune the feedforward gains. Experimental results show that the proposed estimation method has improved the estimation bandwidth over the baseline MRAS and EKF methods by 20 times, and a 0.01 s rise-time was demonstrated in the speed closed-loop step response. Spring Topical Meeting Precision Mechatronic System Design and Control This work may not be copied or reproduced in whole or in part for any commercial purpose. Permission to copy in whole or in part without payment of fee is granted for nonprofit educational and research purposes provided that all such whole or partial copies include the following: a notice that such copying is by permission of Mitsubishi Electric Research Laboratories, Inc.; an acknowledgment of the authors and individual contributions to the work; and all applicable portions of the copyright notice. Copying, reproduction, or republishing for any other purpose shall require a license with payment of fee to Mitsubishi Electric Research Laboratories, Inc. All rights reserved. Copyright c © Mitsubishi Electric Research Laboratories, Inc., 2016 201 Broadway, Cambridge, Massachusetts 02139 ITERATIVE TUNING FEEDFORWARD SPEED ESTIMATOR FOR SENSORLESS INDUCTION MOTORS Lei Zhou, Yebin Wang, David L. Trumper Dept. of Mechanical Engineering Massachusetts Institute of Technology Cambridge, MA, USA Mitsubishi Electric Research Laboratories Cambridge, MA, USA INTRODUCTION For speed sensorless induction motors under field-oriented control (FOC), where the motor speed and angle are not measured, the speed control tracking bandwidth is mainly limited by the convergence rate of the state estimator. Prevailing speed-sensorless induction motors suffer significant performance degradation from removing the encoder, which limits their applications to fields requiring low or medium performance. This paper studies a new estimation approach for induction motors, aiming at improving improving the estimation bandwidth of sensorless induction motors, and thus enabling them for higher bandwidth drives. In the speed-sensorless estimation for induction motors, the classic model reference adaptive system (MRAS) approach [1, 2, 3, 4] was initially studied and remains appealing until today. Although the design was simple, this method often suffers from slow converging due to the adaptive estimation. Through the years, numerous estimation methods have been studied for induction motors, such as sliding mode observer [5, 6], extended Kalman filter (EKF) methods [7, 8], moving horizon estimation (MHE) methods [9], etc. In these designs the rotor’s mechanical equition is often not included in the estimator model, based on the assumption that the mechanical dynamics is slow compared to the electrical dynamics. This assumption can significantly simplify the estimator design and allow the estimation to proceed without knowing the motor’s mechanical parameters, but often results in slow transient. In this work, we propose a new induction motor state estimation method with the rotor’s mechanical dynamics included, targeting at improving the speed estimation convergence rate and thus improve the speed tracking bandwidth of the sensorless induction motor. In the proposed estimator design, the estimations of rotor flux and the rotor speed are separated into two sequential steps, where the flux estimation is achieved by a linear filter, and the speed estimation uses a combination of feedforward and feedback. In order to address the difficulty of unknown rotor inertia and load torque, an iterative tuning method was used to automatically tune the feedforward gains. Experimental results show that the proposed estimation method has improved the estimation bandwidth over the baseline MRAS and EKF methods by 20 times, and a 0.01 s rise-time was demonstrated in the speed closed-loop step response. PROBLEM FORMULATION Assuming symmetric three phase excitation and linear magnetic circuit, the induction motor model in the stationary two-phase frame is given by i̇ds = −γids + ω1iqs + αβψdr + βψqrω + uds/σ i̇qs = −ω1ids − γiqs − βψdrω + αβψqr + uqs/σ ψ̇dr = αLmids − αψdr + (ω1 − pω)ψqr ψ̇qr = αLmiqs − (ω1 − pω)ψdr − αψqr