用PMU实测数据辨识同步发电机参数的关键问题

研究了参数的可辨识性分析、扰动类型的选择、相量测量单元量测数据的滤波和实测功角恒定偏差对同步发电机参数辨识的影响等问题。提出以H阵的条件数作为参数的可辨识性指标,用以定量描述参数的辨识条件,以同时考虑模型结构和扰动深度等因素对参数可辨识性的影响。从理论上分析了不对称扰动数据的正序分量可用于同步发电机的参数辨识,并用实测数据进行了验证。根据对发电机动态特性的激发程度,合理地选择扰动类型;从电子信息科学领域引入一种零相移的数字滤波方法,解决了传统滤波器给量测信号引入相移的问题。总结了现有的功角测量方法,指出直接法测得的功角与真实功角可能存在恒定偏差,并研究了这一偏差对参数辨识结果的影响。实际算例验证了所述方法的有效性。

Key Issues of Synchronous Generator Parameter Identification Using PMU Measurements

The key issues of phase measurement unit(PMU) based parameter identification are studied,including identifiability analysis of parameters,selection of disturbance types,measurement data filtering of PMU,and impacts of permanent bias in power angle measurement on parameter identification.Firstly,condition number of matrix H is taken as the index which can quantitatively describe identification condition.The index can consider the influence on identifiability brought by both model structure and degree of disturbance.Secondly,the fact that positive sequence component derived from asymmetric disturbance in power system can be used for generator parameter identification,is verified by both theory analysis and numeric test using measured disturbance.Some suggestions for disturbance selection are given according to the depth of fault.Then,a kind of zero phase-shift filtering method is introduced to overcome phase shift brought by traditional digital filter.Finally,power angle measurement methods are summarized,and the effect of permanent bias exists in measured power angle to identification result is analyzed.Test results of some actual case studies demonstrate the validity of the proposed method.