Optimal currents based on Adalines to control a Permanent Magnet Synchronous Machine

This paper describes a way to find optimum currents to control a permanent magnet synchronous machine (PMSM) using neural methods. Here the aim is to minimize torque and speed undulations but the proposed method enter in the global problematic of current compensation. Many excellent authors have already written articles on this matter, this one adds a new contribution with the advantage of being flexible toward the process to control. The neural algorithm has been implemented and the effectiveness of the adaptive torque ripple compensation demonstrated by experiments.