Study On IM Sensorless Control System Based On Novel Kalman Filter

In the field of modern AC variable speed, vector control technology performance,with its simple and reliable, has been widely used to control a variety ofhigh-performance AC motor. With the development of manufacturing technology andpower electronics technology, the induction motor (IM) performance and efficiencyhave been improved, the volume is smaller. IM applications more widespread, itshigh-performance speed sensorless control is also widespread concern. Many withoutspeed sensor control method, the Kalman filter with its dynamic performance, theadvantages are not affected parameter changes, etc., can achieve a good IM sensorlesscontrol. However, the traditional extended Kalman filter (EKF) method in the linearsystem of equations of nonlinear processing, error exists on algorithms result inprecise enough. In practice, therefore, the need for conventional Kalman filter isfurther improved. Aiming at the shortcomings of traditional EKF method, Kalmanfiltering method for new in-depth study. Its simulation, estimation accuracy issignificantly improved verification system, and achieved good results.First, the control methods used IM’s brief. IM vector control system using theestablished model of rotor flux orientation method based simulation to verify thecorrectness and validity of the model.Secondly, the traditional extended Kalman filter method was introduced, builtvoltage reconstruction module, establish the feasibility of the EKF-based IM speedsensor control system and its simulation studies show that the method in Matlabsoftware. EKF estimation method for low accuracy problem, the new Kalman filteringmethod conducted in-depth research, namely through the unscented transform (UT) toachieve unscented Kalman filter (UKF).Finally, the UKF speed sensorless control of IM system model based on twomethods by simulation analysis, results show that the new algorithm can improve thespeed of the estimated effect of the system, and change the motor parameters toinvestigate the robustness of UKF.