Investigation On Sensorless Control Of Directly-driven PMSG For Wind Turbines Based On The On-line Parameter Identification

With the growing of the environmental problem, the wind power generation is considered to be the most useful and effective energy. At present, permanent magnet synchronous generator(PMSG) has obtained more application in the variable speed constant frequency(VSCF) wind power generation system, which has advantages of simpler structure, higher generation efficiency and more reliable operation performance, compared with the doubly-fed induction generator(DFIG). Generally, to implement the maximum power point tracking(MPPT) control and keep stable generation performance, it is necessary to detect rotor position and speed to apply vector control for PMSG. However, additional mechanical position sensor not only increases the cost and complexity of the system, but also impairs the reliability. Therefore, it is important to investigate the sensorless control strategy of VSCF performance of PMSG with the consideration of the influence on the PMSG control due to the generator parameter variation.Firstly, the stator voltage model of PMSG in the estimated synchronous reference has been proposed. Then, the mathematical model of rotor position error estimation has been deduced. Secondly, the theory of phase-locked loop(PLL) has been introduced and the PLL controller has been designed to estimate the rotor position and rotor speed. Simulation result and experiment result both show that the PLL controller can obtain good performance.The sensorless operation performance may be deteriorated while PMSG parameters used to the position detecting model vary with operation conditions. Therefore, it is necessary to consider the on-line parameter identification of PMSG to improve the accuracy of the estimated rotor speed and position. The theory of recursive least-squares(RLS) method has been introduced at first. And then, the mathematical model of parameter identification based on RLS has been proposed. At last, experiment result demonstrates that RLS can be utilized in parameter identification.In the end of the thesis, the design and development of a 3.8kW rated hardware system have been proposed. And the experimental results will show that the estimated PMSG parameter can improve the performance of PLL controller.