电力系统关联性参数的机理与辨识

电力系统动态仿真中，关联性参数的不同组合可以得到相同的仿真结果，从而导致根据实际量测辨识得到的关联性参数不唯一，辨识结果有可能偏离真实值。因此，需要对关联性参数的形成机理与辨识方法进行研究。针对参数在数值上不可辨识的现象，以传统结构关联性参数为基础，提出数值关联性参数的问题，并将两者推广为广义关联性参数，其基本特征是轨迹灵敏度线性相关。最后根据轨迹灵敏度线性相关的特点，提出利用选主列QR 分解法与轨迹灵敏度拟合法来进行广义关联性参数识别的方法。算例结果验证了该方法的有效性。

Mechanism of Correlation Parameters of Power System and Related Identification

In power system simulation,different combination of correlation parameters can create same simulation results and consequently cause non-unique correlation parameters based on practical measuring identification,meanwhile,identifi-cation results may deviate from true values. Therefore,it is necessary to study form mechanism of correalation parameters and identification method. Aiming at non-identifiability of parameters in regards to numerical values,taking traditional structural correlation parameters as basis,problem of numerical correlation parameters were proposed. Meanwhile,these two kinds of correlation parameters were extended as generalized correlation parameters of which the basic feature was line-ar correlation of trace sensitivity. According to characteristic of linear correlation of trace sensitivity,a method for identif-ying generalized correlation parameters by using QR decomposition with column pivoting method and trace sensitivity matc-hing method was proposed. Example result verified effectiveness of this method.