HYBRID VECTOR CONTROL FOR INDUCTION MOTOR

Abstract

In this thesis work, first the dynamic modeling of the induction motor has been described because per-phase equivalent circuit analysis it is not appropriate to predict dynamic performance of the motor. In order to understand and analyze vector control of induction motors the dynamic model is necessary. Machine is modeled in stator reference frame since it easier to describe the characteristics of induction motor. Also the machine is modeled in relatively simple terms by using the concept of space vectors and d-q variables. In the next part of the thesis, general principle and various concepts of vector control have been described. Then direct vector control (DVC) has been simulated and analyzed. For the rotor flux position angle estimation, here motor terminal voltages (vim , vq3) and current has been used and rotor flux based calculations are made. Machine is modeled in stator reference frame using state space vectors for viewing actual behavior of inputs and outputs. In direct vector control, only inverse Clark's and Park's transformation are required because torque current and flux current commands are set by using PI controllers. But direct vector suffers from low speed and zero speed problems. Because at these speeds stator voltages, currents and frequency becomes very low and it becomes difficult to sense these quantities. Moreover parameter variation effect of resistance R, and inductances L,j,Lrr and Lm tends to reduce the accuracy of estimated signals. In the next part of the thesis indirect vector control scheme has been simulated and analyzed. This scheme uses rotor speed signal for the estimation of rotor flux position estimation in whole speed range. Here also the machine is modeled in stator reference frame for the same reason as stated above. Since the commanded torque producing and flux producing current are DC signals, so the stator three phase currents are transformed into synchronous frame dq-currents. For this Clark's and Park's are transformations are used. Again for PWM control the actual dq-current that corresponds to input variables (wref and TT) are converted into stationary reference frame by using inverse Clark's and Park's transformations. Parameter variation effects of rotor resistance R, has also been analyzed III both in open loop and closed loop and a method called model reference adaptive control has been used for compensation of rotor resistance mismatch. Since indirect vector control scheme uses rotor speed signal so speed sensing encoders becomes costly and more complex at higher speeds.