基于鲁棒容积卡尔曼滤波器的发电机动态状态估计  

Dynamic State Estimator for Synchronous Machines Based on Robust Cubature Kalman Filter

作　　者：毕天姝 陈亮 薛安成 杨奇逊 

机构地区：[1]新能源电力系统国家重点实验室(华北电力大学),北京102206

出　　处：《电工技术学报》2016年第4期163-169,共7页Transactions of China Electrotechnical Society

基　　金：国家重点基础研究发展计划(973计划)(2012CB215206);国家自然科学基金(51222703);高等学校博士学科点专项科研基金(20120036110009);"111"计划(B08013)资助项目

摘　　要：同步相量测量单元(PMU)能够对电力系统动态过程中发电机功角进行直接量测。然而,坏数据有可能导致状态估计准确度下降甚至失效。提出了一种基于鲁棒性容积卡尔曼滤波(CKF)的机电暂态过程发电机动态状态估计方法。在CKF中构造时变多维观测噪声尺度因子,根据量测新息对PMU量测误差进行调整,使得量测量能够对状态量预报值进行准确修正。给出了时变多维观测噪声尺度因子的具体构造方法。针对滤波增益求逆发生奇异的问题,提出解决方案,对鲁棒CKF动态状态估计过程进行说明。仿真结果表明该方法能够有效抑制量测坏数据对动态状态估计的影响。Phasor measurement unit （PMU） can measure the rotor angle of synchronous machine in power system dynamic process. However, the bad data may decrease the accuracy of state estimations, even lead to the failure of the estimator. Based on the robust cubature Kalman filter （CKF）, a novel dynamic state estimator for synchronous machine in the electromechanical transient process is proposed. A time-varying multi-dimensional scale factor is introduced into CKF. The PMU measure- ment covariance can be adjusted according to the innovation. As a result, the PMU measurements will correct the state predictions precisely. The formulation of the scale factor is clarified, and the method for dealing with the problem of the gain matrix singularity is addressed. The detailed process of dynamic state estimation based on robust CKF is given. The simulation results show that the method can prevent the influence of bad data on the precision of the dynamic state estimation.

关 键 词：机电暂态 动态状态估计 容积卡尔曼滤波 鲁棒性 

分 类 号：TM711[电气工程—电力系统及自动化]