Optimal configuration of dynamic VAR compensators considering uncertainty and correlation

The transient voltage stability of modern power systems is challenged by the increasing uncertainty on source-load dual sides. To cope with it, a robust optimal configuration method for dynamic VAR compensators (DVCs) is proposed. First, a series of power-flow scenarios are generated with uncertainty and correlation considered, where nonparametric kernel density estimation is used to predict the marginal distribution of wind speeds and combined Copula function is employed to characterize correlations between nearby wind farms. Then, the variation-of-information indicator is introduced to assess the similarity among power-flow (PF) scenarios, and representative PF (RPF) scenarios are screened with the help of Spectral clustering algorithm. Finally, locating and sizing of DVCs for RPF scenarios are achieved. By synthesizing the compensation schemes for RPF scenarios, the robust configuration scheme is suggested. Simulation studies in the modified New England 10-machine 39-bus system show the proposed method can ensure transient voltage security of the studied power system under source-load uncertainty.