Stator resistance identification of PMSM

Abstract: The following paper describes an online identification of a stator resistance of a permanent magnet synchronous motor. The resistance is identified by two methods. The first one is a frequency analysis based on an injection of a harmonic signal to the control. This method utilizes knowledge of synchronous motor schema in d-q coordinates and it is based on a mutual signals correlation. The second one-Gauss-Newton method utilizes Hammerstein model. In this method a loss function is created from this model, where the motor is divided into a linear dynamics and static nonlinearity. Both algorithms were tested on a real motor through a real time platform D-space.