RNN Based Rotor Flux and Speed Estimation of Induction Motor

Abstract

Speed control of induction motor can be obtained by closed loop system which require speed sensor. Speed sensor system is less effective for wide plant system, because the sensor location is too far from the main control system and measurement result is less accurate. This paper presents the development of speed sensorless field oriented control (FOC) of induction motor by using the rotor flux and speed observers. The observers only required the stator voltage and current of induction motor to obtain the rotor flux and speed estimation. The observers based on recurrent neural network (RNN) methods are implemented. Finally, the effectiveness of the proposed method is verified by simulation. Simulation results show that RNN observer can produce well the rotor flux and speed estimation. MSE values of the rotor flux estimation are between 0.000087 and 0.000264, whereas MSE values of the speed estimation are between 43.0552 and 156.0798. Keywords: field oriented control, induction motor, observer, and recurrent neural network. DOI: http://dx.doi.org/10.11591/ijpeds.v1i1.65