Direct Torque Control Method using Kalman Filter

Abstract

In classical control of Brushless DC motors, flux distribution is assumed trapezoidal and fed current is controlled rectangular to obtain a aspire to constant torque. However, in reality this assumption may not be correct due to variance of magnetic materials and design trade-offs. These factors together with current controller limitation, can lead to an unwanted torque- ripple. Hence torque - ripple has been the main issue of the servo derive systems, in which the speed fluctuation, vibration and acoustic noise should be diminished. This torque - ripple also degrade the precision of the parameters of brushless de motors. To diminishing the torque ripples, we are executing DTC-SVM method and kalman filter. DTC SVM. Different from the traditional switching table in STDTC, which is composed of a defined number of voltage vectors with fixed amplitudes and positions, SVM can develop an arbitrary voltage vector with any amplitude and length within its linear range; hence, DTC-SVM can govern the torque and flux more accurately and moderately with fixed switching frequency. Another merit of using SVM is that the sampling frequency does not need to be as high as that in traditional DTC. The key point of DTC-SVM is how to obtain the commanding voltage vector. Kalman filter that can be considered an amplification of the Wiener filtering concept. The Kalman filter has an objective the minimization of the estimation square error of a unstationary signal buried in noise these all cannot be precisely minimize the ripples. They all can be minimizing the ripples up to 60%. KF is used based on the estimation of the previous existents. It can be resolved by step-by-step process we can get ripples up to 40%. By using the three methods, we can minimize the torque ripples up to 20%. We cannot achieve 100% torque ripple minimization. The simulation results validate the effective ness of the proposed scheme.