This work relates to the study of direct torque control is applied for salient-pole double star synchronous machine. The machine is supplied by two three-level inverters. Two control approaches using fuzzy logic DTC, and neural network DTC are proposed and compared. The validity of the proposed controls scheme is verified by simulation tests of a double star synchronous machine. The stator flux, torque, and speed are determined and compared in the above techniques. For achieving high performances control of the multiphase drive, the proposed controls method needs accurate information about rotor position and rotor speed. To this end, they are estimated by using Luenberger observer.

Keywords: Double star synchronous machine, Multilevel inverter, Direct torque control (DTC), Fuzzy logic control, Neural network, Luenberger observer