ANN FOR SOLVING THE HARMONIC LOAD FLOW IN ELECTRIC POWER SYSTEMS WITH DG

摘要

In this paper, the use of artificial neural networks (ANN) is proposed for solving the impact of distributed generation (DG) on electric power systems (EPS). Developed models of ANN evaluate the steady state of EPS and are a fundamental for planning, operation and control of modern power systems with DG. The mathematical model of the harmonic power flow comprises a set of non-linear algebraic equations traditionally solved with some iteration method. Variable consumption in the system and variable production of DG must be taken into account of power flow. In order to take into account this fact, we used a Monte Carlo simulation (MCS) to create different situations in the EPS. Using MCS we got a database that is essential for the creation of ANN. Outputs of our database are impacts of DG: active power losses, voltage drops, maximum of voltage drops and total harmonic distortion (THD). The proposed ANN methodology has been successfully tested using the IEEE-33 bus system without DG. Then the same system with DG in different nodes is used for testing created ANN models.