Unscented Kalman Filter based Mechanical Parameter Estimation of Wind Power Systems

Abstract

A variable speed wind power generation system consists a number of controllers. The design of these controllers needs true information of the system parameters for their coordinated control. However, sometimes the exact parameters of the system may not be present which may result the malfunctioning of the employed controllers. Further, the parameters also change with the time and it may also be a reason of mal-operation of the controllers. Thus, there is a need of estimation of the system parameters which gives the exact information of the system parameters and accordingly the operator can update the controller to prevent the mal-functioning. The objective of this work includes online estimation of various mechanical parameters, especially stiffness coefficient and damping constant of the turbine-generator coupling shaft, since the information of the electrical parameters can be found by various simplified tests. Unscented Kalman Filter (UKF) has been used for parameter-estimation with the help of terminal measurements which include generator speed, active and reactive power. The DFIG based wind power generation scheme is used for the study wherein the estimation of the parameters has been performed in various wind speed environments.