基于鲁棒容积卡尔曼滤波器的发电机动态状态估计
发布时间：2019-07-09 06:36
【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。
[Abstract]:The synchronous phasor measuring unit (PMU) can measure the power angle of generator directly in the dynamic process of power system. However, bad data may lead to the decline or even failure of state estimation accuracy. A dynamic state estimation method for electromechanical transient process generator based on robust volume Kalman filter (CKF) is proposed. The time-varying multi-dimensional observation noise scale factor is constructed in CKF, and the PMU measurement error is adjusted according to the measurement innovation, so that the state quantity prediction value can be accurately corrected. The concrete construction method of time-varying multi-dimensional observation noise scale factor is given. In order to solve the singular problem of inverse filtering gain, a solution is proposed, and the process of robust CKF dynamic state estimation is explained. The simulation results show that the method can effectively suppress the influence of bad data on dynamic state estimation.
【作者单位】： 新能源电力系统国家重点实验室(华北电力大学);
【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
【分类号】：TN713;TM31
，
本文编号：2511933
[Abstract]:The synchronous phasor measuring unit (PMU) can measure the power angle of generator directly in the dynamic process of power system. However, bad data may lead to the decline or even failure of state estimation accuracy. A dynamic state estimation method for electromechanical transient process generator based on robust volume Kalman filter (CKF) is proposed. The time-varying multi-dimensional observation noise scale factor is constructed in CKF, and the PMU measurement error is adjusted according to the measurement innovation, so that the state quantity prediction value can be accurately corrected. The concrete construction method of time-varying multi-dimensional observation noise scale factor is given. In order to solve the singular problem of inverse filtering gain, a solution is proposed, and the process of robust CKF dynamic state estimation is explained. The simulation results show that the method can effectively suppress the influence of bad data on dynamic state estimation.
【作者单位】： 新能源电力系统国家重点实验室(华北电力大学);
【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
【分类号】：TN713;TM31
，
本文编号：2511933
本文链接：https://www.wllwen.com/kejilunwen/dianzigongchenglunwen/2511933.html
