摘要
全国电网互联使电力系统的规模越来越大、运行状态越来越多变,研究基于广域测量数据的系统模型在线辨识方法成为电力系统在线分析及广域控制的重要理论问题。考虑初始状态对系统输出的影响,提出一种电力系统传递函数的通用普罗尼(Prony)辨识方法,同时给出激励信号和数据采样参数的选取原则。该算法允许待辨识系统初始状态非零,可采用任意类型的激励信号,并能综合处理多时段数据。辨识4机2区域系统和新英格兰系统的传递函数模型,仿真分析结果表明,通用Prony辨识方法能准确辨识初始状态和传递函数,计算速度快且几乎不随系统规模而变化,辨识方法具备在线应用的潜力。
With interconnection of power grid, power system size becomes larger, and operation condition becomes more variable. Online identification of system model based on wide-area measurement data has become an important academic question for online analysis and wide-area control of power system. Considering the effect of initial states on system output, a general Prony identification algorithm of power system transfer function is brought forward, with selection principles of pumping signals and sampling parameters presented afterward. This algorithm permits the appearance of non-zero initial states and any type of pumping signals in the system to be identified, and could deal with multiple time-intervals situation. Identification of transfer function is carried out on two-area four-machine power system and new England test power system （NETPS）. Simulation and analysis results show that this algorithm identifies initial states and transfer function accurately, moreover computation-time for identification is short and almost identical when system size changes. The general Prony identification algorithm is promising for online appfications.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2008年第28期41-46,共6页
Proceedings of the CSEE
基金
国家自然科学基金项目(50507018
50595414)
国家"十一.五"重点科技攻关项目(2006BAA02A01)
教育部科学研究重大项目(305008)~~
关键词
广域测量系统
电力系统模型
系统辨识
PRONY算法
传递函数
wide-area measurement system
power system model
system identification
Prony algorithm
transfer function
