Torque Ripple Suppression Control Using a Parallel Resonant Controller and an ANN in the Position and Acceleration Sensorless Control of Stepping Motors

抄録

In 2-phase HB-type stepping motors, since the phase A and B induced voltages contain 3rd and 5th order components, the dq-axis induced voltage contains 4th-order component. Therefore, the dq-axis magnetic flux includes a 4th-order component, and the output torque also includes a 4th-order component. To suppress this torque ripple, conventional methods include ripple suppression using position and acceleration sensors and manual adjustment of the q-axis current commands. However, these methods increase cost, reliability, and labor. Therefore, this paper reports a new method of torque ripple suppression using position and acceleration sensorless control by improving the current response using a resonant controller and generating q-axis current command values using an artificial neural network (ANN), which is a type of AI model. In this paper, we also confirm the effectiveness of suppressing torque ripple in the mid-to high-speed range through actual machine verifications.