不确定量测下发电机动态状态估计性能分析

Performance Analysis of Generator Dynamic State Estimation under Uncertain Measurement

相量测量单元(PMU)中随机误差不可避免,在实际电网系统中PMU量测数据可能出现延时、重新排序甚至丢失等不确定情况。为准确估计电力系统机电暂态过程中的状态信息,首先建立量测丢失下的发电机动态状态估计模型;然后在某实际电网系统算例中分别采用无迹混合滤波(UMF)、粒子滤波(PF)和所提出的改进粒子滤波(IPF)3种算法对发电机动态状态估计模型进行了仿真试验。仿真结果表明:在不确定量测系统下,改进的IPF算法的滤波性能和抗差性能优于UMF与PF算法,更适用于不确定量测下发电机动态状态估计。

Random errors are unavoidable in phasor measurement unit(PMU),and the PMU measurement data may be uncertain in actual power system,such as delay,reordering or even missing.In order to accurately estimate the state information in the electromechanical transient process of power system,a generator dynamic state estimation model is firstly established under the missing measurement;And then,the model is simulated in an actual power system using the unscented mixture filter(UMF),particle filtering(PF)and the improved particle filtering(IPF)proposed in this paper respectively.The results show that,under uncertain measurement,the proposed IPF is superior to UMF and PF in filtering performance and robust performance,and more applicable to generator dynamic state estimation.