Multivariable Reconstruction for Permanent Magnet Synchronous Motor with Input Filter Based on Extended State Variable

Active damping is important for using input filters to suppress resonance in permanent magnet synchronous motors (PMSM), but additional sensors are often required. A new multivariate reconstruction (MVR) method is proposed, which is based on the current of inverter and can estimate the current and voltage of the motor with high accuracy and strong robustness, thereby achieving active damping without additional current or voltage sensors. The accuracy and robustness of the proposed MVR in damping systems were tested. Experiment was carried out to test the effectiveness of MVR.