Sensorless speed control and initial rotor position estimation of an interior permanent magnet synchronous motor drive

In this paper, a new approach to a sensorless speed control and initial rotor position estimation for interior permanent magnet synchronous motor (IPMSM) drive is presented. At rotating condition, speed and rotor position estimation of IPMSM drive are obtained through an extended Kalman filter (EKF) algorithm, by simply measurement of the stator line voltages and currents. The main difficulty in developing an EKF for IPMSM is caused by the complexity of the dynamic model expressed in the stationary coordinate system. This model is more complicated than that of the surface PMSM, because of the asymmetry of the magnetic circuit. The starting procedure is a problem under sensorless drives, because no information is available before starting. The initial rotor position is estimated by the current response of the voltage pulse, which is intermittently applied to the motor at standstill. To illustrate our work, we present experimental results for an IPMSM obtained on a floating-point digital signal processor (DSP) TMS320C31/40 MHz based control system.