A Low-complexity Encoderless Model Predictive Current Control using Luenberger Observer for Induction Machine Drives

This work would like proposes a low-complexity encoderless model predictive current control (MPCC) using Luenberger observer (LO) for induction machine (IM) drives. The proposed low-complexity MPCC optimizes the current error in the designed cost function without an exhaustive search. However, the information of angular speed is still collected by an encoder, which brings in an increased system maintenance and reduced reliability. Specifically, the flux information is estimated by the model of control plant, which is very sensitive to the variations of system parameters. To tackle the aforementioned issues, a Luenberger observer is proposed for estimating both the angular speed and model parameters in the MPCC methodology. The experimental implementation is conducted on a 2.2 kW IM testbench, which confirms a decent performance in a series of test scenarios.