A self organizing intelligent controller for speed and torque control of a PMSM

In this paper, an intelligent self-organizing method for speed and torque control of a PMSM is proposed. The mechanical dynamic of the machine is supposed comparable to the electrical one. Two cases are studied: speed and torque control in presence of a rotor position sensor, and mechanical sensorless torque control. The obtained controllers are two dynamic recurrent neural networks with an almost optimized structure, trained by an original hybrid training method. Simulation results show that the closed-loop system responses are satisfactory.