Automatic power factor correction based on Pearson similarity for distribution networks.

In this paper, an online algorithm of automatic power factor correction (APFC) for distribution power grids is presented. The algorithm depends on Pearson similarity of voltage and current signals, which can easily be measured locally using voltage and current transformers, respectively. The proposed method not only evaluates the original power factor (PF), but also functions the APFC more quickly. In this work, the compensation index and reactive power compensation can be identified using new mathematical formulas based on the cross-correlation coefficients calculated between the phase voltage and current signals. Moreover, the coefficients are useful for detecting any abnormal condition. The method performance is examined on a large-scale network with real parameters' data, which is simulated using the alternative transient program (ATP) environment, and the proposed algorithm is developed in MATLAB package. Numerous simulation results confirm the APFC relay feasibility, and the unity PF can be achieved. Besides, the proposed algorithm is characterized by its high operating speed, safety and accuracy for PF compensation, where its performance accuracy is greater than 99.7%. The approach offers novel operating/tripping curves based on the correlation coefficients for both power factor correction (PFC) and protection functions. Therefore, it is described with high level of functionality integration, and its security and sensitivity are controllable easily using the correlation settings. [ABSTRACT FROM AUTHOR]