Intelligent controllers for sensorless induction motor drives

Abstract

For high performance industrial applications, the sensorless vector controlled induction motor has been the most frequently used one for the past two decades. Such strategy reduces the cost, size of the drive system and maintenance requirements. Reliability and robustness are increased because of the sensorless control. Sensorless drive performance is good during medium and high speed operations but deteriorated near zero and low speed regions. This is due to the low or even zero rotor induced EMF at zero stator frequency. Hence, the speed estimation fails at low and zero speed operating regions. Much recent research works have been proposed to improve the drive performance near zero stator frequency. Among several model based speed estimation techniques, Model Reference Adaptive Systems (MRAS) are famous due to their less computational effort and simplicity. The three components of MRAS are Reference Model (RM), adaptive model and adaptation mechanism. The stator voltage equations in the stationary reference frame represent RM or Voltage Model (VM) and these equations are independent of the rotor speed. The Current Model (CM) represents newlinethe adaptive model whose equations are dependent on rotor speed. The adaptation mechanism is designed using the conventional Proportional Integral (PI) controller to generate the estimated speed. Popov s stability criterion is used for the design of the adaptation mechanism. Rotor flux MRAS (RFMRAS) scheme proposed by Schauder is used in this work to improve the system performance. newline newline