Flux Linkage Estimation for PMASynRM MTPA Control: Comparing ANN and LUT Techniques

This paper presents a comprehensive investigation into the critical role of accurate flux linkage estimation for achieving optimal Maximum Torque Per Ampere (MTPA) control in Permanent Magnet Assisted Synchronous Reluctance Motors (PMASynRMs). The study compares two offline flux linkage estimation methods: an Artificial Neural Network (ANN), employed for online computation, and a Look-Up Table (LUT), utilizing precomputed values. The research assesses the accuracy of each method in estimating the nonlinear flux linkage of the motor, which is essential for precise control, particularly given the complex magnetic characteristics of PMASynRMs, which exhibit direct and cross-saturation effects. Simulation and experimental results presented in this work demonstrate that the ANN estimator more effectively captures the complex, nonlinear relationships between current and flux linkage in PMASynRMs, outperforming the LUT estimator, particularly in regions of high magnetic saturation