Elbow-shaped criterion-based coherency machine group clustering identification method

The invention provides an elbow-shaped criterion-based coherent generator group clustering identification method, which comprises the steps of reading BPA stable data to form a generator rotor swing angle difference matrix; selecting a plurality of clustering centers by adopting a maximum and minimum distance method; estimating the optimal clustering number according to a GSA-based elbow criterion; and performing clustering analysis according to the optimal clustering number to obtain a plurality of coherency machine groups. According to the invention, a GSA-based elbow criterion is adopted tocarry out homology discrimination on the generator set; the optimal clustering number can be automatically estimated, the automatic judgment result is accurate and conforms to the reality, the localinfluence caused by large offset of individual points is considered, global statistics of oscillation energy is also considered, the adaptive capacity is high, meanwhile, the clustering center is selected by adopting the maxim