Identification and Online-correction of Periodic Motor Torque Fluctuations on Electromechanical Axes

Electromechanical axes generate high-precision motion profiles and positioning sequences in a wide range of industrial applications. The utilization of the drive internal signals for monitoring purposes, e.g. for estimating the axis load during a production process, is gaining more and more attention. One major drawback is the superimposition of the load-proportional motor torque by further operation-related disturbance effects. The paper presents a novel approach for the identification and correction of periodic disturbance fluctuations in the motor torque signal. Therefore, a novel method for automatic offline determination of the relevant spectral components using the least squares method is applied. The identified magnitude and phase values are stored in speed-related lookup tables. During normal operation, the sum of the individual sinoids is subtracted from the measured motor torque. The paper provides the overall description of the methodology and its functional verification on a single-axis drive test rig equipped with an industrial control.