Reduced order extended kalman filter based parameter estimations for speed-sensored induction motor drive

In this study, for a high performance speed-sensored control in fieldweakening region of induction motors, a novel reduced order extendedkalman filter based estimator is designed and tested by experimentsperformed in simulation and real-time experiments. the proposedreduced order extended kalman filter estimates the stator stationaryaxis components of rotor fluxes, which are required for vector controlsystem. in addition to the flux estimations, rotor resistance and mutualinductance whose values vary according to operating conditions ofinduction motors are estimated simultaneously in a wide speed rangeincluding zero-speed and field-weakening. the simulation basedestimation results associated with proposed estimation algorithm arequite satisfactory under challenging variations of rotor angularvelocity, load torque, rotor resistance, and mutual inductance.therefore, the performance of the direct vector controlled inductionmotor drive using the proposed estimation algorithm also becomesquite well. moreover, the estimation results in real-time experimentsalso confirm the performances of the proposed estimator.