Harmonic Voltage Estimation of Distribution Networks Using Inverse PCA

This paper presents a new methodology to estimate harmonic distortion in a distribution system, based on measurements of a limited number of given buses. The algorithm utilizes Principal Component Analysis (PCA). First the PCA of buses is detected off-line from their historical data. Next, the interrelations between the voltages are developed using Inverse PCA method. Finally, through incorporation of a certain finite set of real-time data measurements, the voltages of all network buses in real-time are estimated with a desired degree of accuracy via the Inverse PCA method (IPCA). The algorithm is applied to an existing industrial distribution network as a test case. This case study shows good agreement between proposed methods with on site measurements.