An Improved Approach to Anticipative Motion Control for Dynamic Efficiency in Induction Drives

This paper targets energy efficiency in induction motor drives under dynamic operation. We propose an extension of a gradient-based augmented-Lagrangian modelpredictive control algorithm (GRAMPC) that selects, at each instant during transients, the optimal combination of flux- and torque-producing stator-current components to minimize total losses. An anticipative feature is introduced that admits timevarying reference and load trajectories known over the prediction horizon, rather than step changes only. The approach is evaluated against a step-based GRAMPC baseline using MATLAB/Simulink studies and rapid prototyping on a dSPACE platform. Experiments show a significant reduction in specific energy losses during speed ramps, with consistent benefits across tested acceleration rates.