Neural-net based real-time control of capacitors installed on distribution systems

Abstract

An expert system using a two-stage artificial neural network is proposed to control in real time multicap capacitors installed on a distribution system for a nonconforming load profile such that the system losses are minimized. The required input data are directly obtained from online measurements which include the active and reactive line power flows, voltage magnitudes, and the current capacitor settings at certain buses. The optimum control does not involve any iteration procedure; therefore, it is computationally very efficient. Studies on a 30-bus distribution test system show the expert system to have quite satisfactory results.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>