基于先验方差的发电机惯量辨识数据质量评估

发电机惯量是电力系统频率特性分析与在线应用的重要参数。基于发电机正常运行时机端有功功率和频率的类噪声信号可对发电机惯量进行实时辨识。然而实测数据质量存在的缺陷，导致现有算法对实测数据辨识效果较差。为解决该问题，本文以谱分析与系统辨识理论为基础，通过参考系统估计、模型参数方差估计、惯量方差估计三个步骤，建立惯量辨识结果的先验方差统计量，在进行辨识前对类噪声数据段进行评价和筛选，提升了惯量辨识的准确度。基于仿真数据和实测数据的数据评估筛选结果验证了本文提出方法的有效性。结果表明，先验方差较小的数据段，惯量辨识的准确度较高。

Data quality evaluation of generator inertia identification based on prior variance

Generator inertia is an essential parameter in the frequency characteristic analysis and online application of power system. The inertia of generators can be identified in real time based on ambient active power and frequency signals while the generator is in routine operation. However, due to the defects of the quality of real world data, the identifi-cation results of existing algorithms applied to real world data are poor. To solve this problem, based on spectral analysis and system identification theory, through three steps of reference system estimation, model parameter vari-ance estimation and inertia variance estimation, the a priori variance of inertia identification results is established. The segments of ambient data are evaluated and selected before identification, which improves the accuracy of inertia identification. Data evaluation and selection results based on simulation data and field measurements verify the method proposed. The results show that the data segments with smaller a priori variance has higher accuracy of iner-tia identification.