参数辨识依赖于激励信号与观测量的选择。该文根据观测量的模式增量计算,确定系统动态在不同观测量上的可观性,据此选择观测量。根据不同扰动激发出的系统主导动态,确定在该扰动下的重要参数。采用分步辨识思路,先根据电网侧故障辨识双馈风力发电机电气部分参数,再基于输入侧风速变化辨识机械部分各参数。辨识方法采用全局最优位置变异粒子群算法,仿真算例验证了该方法的有效性。
Parameter estimation relies on disturbance signals and measurements selection. The mode increment of measurements is calculated to quantify the observability of system dynamics on each measurement. The connection between the type of disturbance and important parameters is explored. The 2-stage estimation process is provided. The first stage uses a 3-phase fault in power grid to estimate the electric parameters, and the second stage considers a wind speed disturbance in order to estimate the parameters in mechanical component. The estimation results validate the effectiveness of this method.