电力系统戴维南等值参数辨识的不确定模型  

Uncertain Models for Identification of Electric Power System Thevenin Equivalent Parameters

作　　者：廖国栋[1] 王晓茹[1] 

机构地区：[1]西南交通大学电气工程学院,四川省成都市610031

出　　处：《中国电机工程学报》2008年第28期74-79,共6页Proceedings of the CSEE

基　　金：国家自然科学基金项目(50595411;90610026);四川省应用基础研究项目(2006J13-111)~~

摘　　要：考虑到系统的运行方式、故障扰动及非线性负荷的动态行为等因素对戴维南等值参数的影响,在状态空间建立电力系统戴维南等值系统的离散时间不确定模型。融合预报协方差矩阵的自适应调整和状态空间鲁棒性估计,给出辨识戴维南等值参数的新算法。采用基于UDUT分解的测量更新及时间更新算法,使算法具有良好的数值稳定性。在改进的IEEE 50机测试系统上,对提出的辨识算法进行仿真测试。测试结果表明,算法具有好的鲁棒性、数值稳定性和抗干扰能力,其在功率增长条件下的等值参数辨识结果符合PV分析得到的最大功率值结论。Considering the effects of time-varying factors as operation condition change, disturbances and nonlinear load dynamics on Thevenin equivalent parameters in power systems, this paper defined the state-space description of Thevenin equivalent systems with uncertain models. Based on the robust estimation in the state-space and the adaptive modification of prediction covariance matrix, a new algorithm was proposed to track Thevenin equivalent parameters. And the measurement updating and time updating using the U-D factorization were used to guarantee the filtering numerical stability. In a revised IEEE 50 generators test system, the proposed tracking algorithm was tested by some simulations. The simulations show the new algorithm has good robustness, numerical stabilities and powerful capabilities of anti-jamming, and the result of maximum loads based on the Thevenin equivalent parameters tracked by the new filter with the load increasing is accorded with the result of a PV analysis at the testing bus.

关 键 词：电力系统戴维南等值 卡尔曼滤波 鲁棒性估计 UDU^T分解 参数辨识 

分 类 号：TM712[电气工程—电力系统及自动化]