A new formulation IPCA based method for branch-current estimation in distribution networks

This paper introduces a neuro-based branch´s current state estimator for distribution networks in which a limited set of real-time data is available. This estimator employs inverse principle component analysis (IPCA) method. Using the IPCA method leads to some proper models devised so that the distribution network becomes observable and then load estimation will be possible. By determining load consumption in each node and voltage magnitude of the reference bus, load flow program can be run for determining the power flow of all sections, which are called pseudo power measurements. Finally, through incorporation of a certain and finite number of variables including real-time voltage magnitudes in few of buses, current magnitude in the beginning of the feeder and pseudo power flow measurements, the currents of all the network´s branches are estimated in real-time via the IPCA and WLS methods with a desired degree of accuracy. A case study on a typical network is considered at the end of paper for better highlighting of the merits of the proposed method. The experimental results show that the proposed neuro-estimator outperforms the previously cited techniques.