Sensorless Control For Surface Permanent Magnet Synchronous Motor And Detection Of Initial Position

Sensorless control for permanent magnet synchronous motor has got great developments in past ten years. However, very few papers have been published on the surface permanent magnet synchronous motors (SPMSMs). The difficulty with the nonsalient rotor is the equal inductances along both d and q axes. So, the emphasis of this thesis is sensorless control for SPMSM.Firstly, mathematics model of SPMSM is set up. PWM technology based on space voltage vector is presented and at the same time, a PWM scheme for DSP with the voltages of two phase stator coordinate is described. The simulation model of PWM inverter for close-loop control with MATLAB is discussed and results of space voltage PWM (SVPWM) are presented.Secondly, a kind of self-adaptive observer is employed to complete estimations of rotor position and speed for sensorless control. After that, the comparison about performances and characteristics of close-loop control system with and without a mechanical sensor is presented.What's more, a new scheme for rotor initial position estimation, voltage pulse vector method (VPVM) is introduced. The estimation is performed by the nonlinear magnetization characteristics of the stator core caused by the magnet of the rotor. This method is based on the principle that the X-axis current value due to the voltage vector applied to the motor under magnet saturation conditions increases as the voltage vector generated from the inverter approaches the N pole of the rotor. Finally, experimental platforms are set up using DSP to complete the estimation of rotor initial position.