The advance in the wide-area measurement system (WAMS) is driving the power system to the trend of wide-area monitoring and control.The Prony method is usually used for low frequency oscillation online identification....The advance in the wide-area measurement system (WAMS) is driving the power system to the trend of wide-area monitoring and control.The Prony method is usually used for low frequency oscillation online identification.However,the identified amplitude and phase information is not sufficiently used.In this paper,the amplitude is adopted to detect the occurrence of the oscillation and to obtain the mode observability of the sites.The phase is adopted to identify the oscillation generator grouping and to obtain the mode shapes.The time varying characteristics of low frequency oscillations are studied.The behaviors and the characters of low frequency oscillations are displayed by dynamic visual techniques.Demonstrations on the "11.9" low frequency oscillation of the Guizhou Power Grid substantiate the feasibility and the validation of the proposed methods.展开更多
文摘该文将结构学领域用于分析建筑或机械结构振动特性的多参考点复指数(poly-reference complex exponential,PRCE)法引入电力系统,实现低频振荡辨识。该方法基于多通道信号,构造脉冲序列的自回归模型(autoregressive model,AR model),运用LQ分解,求解其自回归系数,再求解由自回归系数构成的矩阵多项式的根,得到频率和阻尼比,进而求得系统的模态矩阵。该文介绍了PRCE法在电力系统低频振荡模式辨识中的原理,16机系统的仿真结果验证了PRCE方法的有效性和正确性;与总体最小二乘–旋转不变技术参数估计(total least squares-estimation of signalparameters via rotational invariance techniques,TLS-ESPRIT)方法、随机子空间(stochastic subspace identification,SSI)方法的对比表明,该方法辨识结果在辨识精度、效率上表现得更好。最后,基于四川电网实测数据验证了PRCE方法的有效性。
基金supported by the National Natural Science Foundation of China (Grant No. 50595413)
文摘The advance in the wide-area measurement system (WAMS) is driving the power system to the trend of wide-area monitoring and control.The Prony method is usually used for low frequency oscillation online identification.However,the identified amplitude and phase information is not sufficiently used.In this paper,the amplitude is adopted to detect the occurrence of the oscillation and to obtain the mode observability of the sites.The phase is adopted to identify the oscillation generator grouping and to obtain the mode shapes.The time varying characteristics of low frequency oscillations are studied.The behaviors and the characters of low frequency oscillations are displayed by dynamic visual techniques.Demonstrations on the "11.9" low frequency oscillation of the Guizhou Power Grid substantiate the feasibility and the validation of the proposed methods.
