扩展卡尔曼滤波在动态负荷参数辨识中应用

Application of extended Kalman filter in parameter identification of dynamic load model

采用扩展卡尔曼滤波算法建立由动态负荷和静态负荷组成的综合负荷数学模型,并列出了其转子运动方程、状态方程和输出方程,其中动态负荷由等值的异步电机表示,静态负荷由恒定导纳并联组成。通过动模试验,取得给定负荷在系统扰动时的电压、电流数据。根据所建立数学模型的输入、输出值,用扩展卡尔曼滤波算法辨识其中的待定参数。参数初值设置为真值的2～7倍,辨识结果误差为2%～3%。分析结果表明,扩展卡尔曼滤波可在短时间内收敛,能正确地辨识出系统参数,且稳定性好。结论表明扩展卡尔曼滤波可以用于电力系统参数辨识,为电力系统状态估计、负荷建模提供了有效方法。

The mathematical model of integrated load composed of dynamic and static loads is established using the extended Kalman filter algorithm and its rotor movement equations,state equations and output equations are listed,in which the dynamic load is expressed by the equivalent asynchronous machine and the static load is expressed by the inductance in parallel with the capacitance. Voltages and currents of a given load during system disturbance are acquired through dynamic simulations. Based on inputs and outputs of the model,the extended Kalman filter is used to identify the unknown parameters. Initial values of those parameters are set at two to seven times of their true values,while the identification error is about 2 % to 3 %. Results show that,the extended Kalman filter algorithm converges in a short time and identifies system parameters correctly with good stability. It provides an effective approach to the state estimation and load modeling of the power system.