Reduction technique of speed ripple using learning control for small brushless motors.

Abstract

As reduction technique of speed ripple of small brushless motors, we developed two new speed control methods, using learning control theory. One is a ripple compensation control method which is canceled by motor torque ripple and compensating torque which is made by calculating the magnitude and phase of the torque ripple of arbitrary frequency in the speed ripple using Fourier transformation. The other is optimum gain control which searchs for the gain giving the minimum speed ripple. We applied them to motor control for audio-video and communications equipment. By applying the ripple compensating control methed to three frequency components, we could reduce speed ripple to 13_??_19 dB compared to values obtained when using no ripple compensating control method and we could cut the speed ripple magnitude in half. By applying the optimum gain control, we could reduce the speed ripple when a filter function was employed to calculate speed ripple. The speed ripple magnitude was reduced by half for these control method also.By applying both system simultaneously, we could reduce speed ripple by one-third compared to that of ordinary systems, and it took 12_??_19 s to finish learning.