Research On The Control Strategy For A Speed-sensorless Induction Motor Based On Model Reference Adaptive System

The main problem of speed-sensorless vector control drive systems of induction motors is bad low speed performance, and it is caused by voltage model and its back electromotive force (back EMF) calculation of motors. To solve this problem, the vector controlled induction-motor drive with a model reference adaptive system (MRAS)-type speed estimator was analyzed through modeling and simulation, and a stator-current-based MRAS type (MRASM-TCC) speed estimator in the synchronous reference frame was developed.Two models were built on the platform, using a rotor-flux-error-based MRAS (MRASF) estimator and a stator-current-based MRAS (MRAS?-?CC) estimator in the stationary reference frame separately. For comparative purposes, the two models were simulated and the MRAS?-?CC estimator was assessed in terms of speed-estimation error compared to the MRASF estimator. Simulation results demonstrate that the MRAS estimator is a right direction to solve the low speed problem, using the motor self as a reference model instead of the voltage model and the stator model together with the current model of the motor as an adjustable model.To solve the problems of the MRAS?-?CC estimator, a stator-current-based MRAS type (MRASM-TCC) speed estimator in the synchronous reference frame was developed. In the estimator, the motor itself was selected as a reference model, and the stator model together with the current model of the rotor flux as an adjustable model. The current model in the synchronous reference frame was used to solve the problems of calculating and controlling the AC variables in the stationary reference frame caused by the noise at the low speed. Furthermore, an error compensator with a real-number gain was added in the stator model to eliminate the error of the flux observation, and the error changed with the disturbance of the rotor resistance parameter. The MRASM-TCC estimator was compared with the MRAS?-?CC estimator by MATLAB in the low speed range. The simulation results show that the MRASM-TCC speed estimator in the synchronous reference frame is accurate. The system has a good dynamic and steady state performance in the low speed range, and also improves the robustness to the changes of motor parameters.