Online coherency analysis of synchronous generators in a power system

In multi-machine power systems, synchronous generators tend to oscillate in several coherent groups, each group being equivalent to a virtual generator. Coherency analysis is fundamental to wide area control of large power systems. Usually, coherency analysis is carried out in an offline mode. However, in response to various events at different operating conditions, the coherent groups may differ. Thus, it is important to develop the analysis to be online. In this paper, K-harmonic means clustering (KHMC) approach is introduced for online analysis. This approach is insensitive to initialization of group centers and is fast. Besides, a second algorithm is developed to automatically determine the optimal number of groups during the online analysis for KHMC. Simulation results are presented using the IEEE 68-bus 16-machine power system. The results indicate that KHMC approach correctly identifies the group centers and assigns each generator to its corresponding group.