This study introduces a new method to calculate the most efficient point of operation of a Permanent Magnet Synchronous Machine (PMSM). The proposed method combines the characteristics of an analytical solution known as the Maximum Torque Per Ampere (MTPA) curve with compensation through an Artificial Neural Network (ANN). By employing this hybrid approach, it is possible to identify and optimize the operating point, even when there is uncertainty in the parameters used in the analytical model.