Model Predictive Control of LC Filter Equipped Surface-Mounted PMSM Drives Using Exact Discretization

Abstract

In this article, a novel method is presented to exactly discretize the state-space model of surface-mounted permanent magnet synchronous machines (PMSMs) that are equipped with an inductive-capacitive (LC) filter to allow the model predictive control of the system. An approach is presented to calculate the matrix exponential and the inverse of the angular frequency dependent state-matrix required for the exact discretization. This enables real-time implementation without using memory extensive lookup tables. The method is compared with the Taylor series up to the third term and bilinear approximations as well. Measurement comparatively verifies the theoretical findings in steady-state at various sampling rates, computational requirements, and prediction accuracy experiments. Furthermore, the robustness of the system to speed reference, load torque, and parameter change transients is also explored.