Research On Sensorless Vector Control Of ACIM For Electric Vehicles

The application of the vector control technology makes the AC system a good dynamic and static performance. As the control theory becomes more and more mature, with its advantages of low cost, high reliability and simple maintenance, AC induction drive system gradually developed into the mainstream of automotive electric drive system. Based on the nonlinear kalman filter theory, this paper concentrates on the study of sensorless vector control of AC induction motor for electric vehicle. Both simulation and experiment are carried to evaluate the performance of the control system.Coordinate transform theory is deduced and then the motor models under different coordinate system are established in turns. The vector control principle is also described in details. To validate the control theory, a vector control system simulation model and an experimental platform of AC induction motor is established. The simulation and experimental results confirm that the system is with a good performance.The paper emphatically analyses the principle and iterative process of the extended kalman filter theory and unscented kalman filter theory. Combined with the state space equations model, speed and flux estimation models for AC induction motor based on extended kalman filter theory and unscented kalman filter theory are established respectively. The iteration steps for state estimation of these two different models are discussed in details. In order to validate the performance, two simulation models in MATLAB/Simulink environment are built. The simulation results demonstrate that two models ensure accurate estimate and both with perfect dynamic and static performance.Combined with theoretical analysis and simulation model, an experimental platform based on the digital motor control development box from TI is achieved. In order to verify the speed and flux estimation effect, this paper design and carry static and dynamic experiments. The experimental results further prove that the designed system has good performance. Considering the influence of rotor resistance, estimate effect is also investigated by changing the rotor resistance. The results verify the system has good adaptability to the change of rotor resistance. From the analysis of the waveforms of speed, rotor fluxs, stator current, system based on the unscented kalman filter theory has faster dynamic response and less dynamic and static estimation errors.In order to further improve the effect of system state estimation, phase voltage distortion problem is analyzed. First, this paper enumerates the causes of this phenomenon and quantitatively calculates the phase voltage error and also deduces the error voltage formula. Then, a self-adaptive compensation strategy focused on compensation time and current symbol is proposed. Combined with the unscented kalman filter theory, the error voltage compensation experiment is carried. From the experimental results, we can see an obvious effect which means the compensation strategy is very effective. After the compensation, the estimated speed and flux of the system are closer to the real value. The whole control precision of the system is improved.