Fourier Neural Networks for real-time harmonic analysis

Nowadays with the technological advances, real time determination of harmonic levels in power systems is a popular research topic. In this paper, to determine harmonics fast and accurately, a new method combining the advantages of Fourier Analysis (FA) and Artificial Neural Networks (ANN) is used. This structure can be defined as Fourier Artificial Neural Network (FANN). The mathematical expressions of the structure are presented in details. Performance of the method was tested with the signal applications containing harmonics and noise in the simulation environment. We found that the method gives faster or more accurate results compared to ANN and FA methods in real time applications.