Adaptive compensation algorithm for flux estimation of PM BLDC motor drives

Abstract

Low-cost Hybrid Electric Vehicle uses a Permanent Magnet Brushless DC Motor with non-sinusoidal back-emf and generally does not use a costly rotary shaft encoder for finding out rotor speed and position accurately. Therefore, it is important that we estimate the rotor speed by some other means. Accurate estimation of flux is one such means. Normally, we estimate flux by integrating the back-emf derived through the voltages and currents of the motor resolved in the stationary reference frame, but in the absence of initial values of flux linkage, the integration produces some DC offset problem in the estimation of flux. Due to this inaccuracy, the estimation of flux linkage angle, speed and torque in a sensor-less drive becomes inaccurate. In this paper, it is shown that the error due to normal integration can be easily eliminated using a New Integration Algorithm suggested for AC systems. The method has been used on PM BLDC motor to eliminate the dc drift in stator flux in achieving correction in the estimation of stator flux angle.