Power Flow Analysis by Numerical Techniques and Artificial Neural Networks

Abstract

This paper discusses different Power Flow Analysis methods. The methods are either Numerical Techniques as the Newton-Raphson (NR) and Gauss-Seidel (GS) methods, or the Artificial Neural Networks (ANNs) method. To solve the power flow problem, we developed Matlab programs using both NR and GS methods. Then we established and trained an artificial neural networks. We found that even though the NR method's algorithm is more complex than GS's algorithm, the NR method is faster and more efficient for a large system. Then, if we have wisely chosen and trained an artificial neural networks to be used for computing voltages magnitudes and voltage phase angles of the Power Flow problem, we can have satisfactory results. Tests are carried out on the Institute of Electrical and Electronics Engineers (IEEE) 14 bus system.