Determining true harmonic contributions of sources using neural network

This paper proposes exact radial basis function neural network (ERBFN) to identify the harmonic sources and their contributions at the point of common coupling without disconnecting the load from the network. The main advantage of the method is that only waveforms of voltages and currents have to be measured. This method is applicable for both single and three phase loads. Comparisons are made with the different types of neural networks such as feed forward back propagation neural network (FFBPN), cascade feed forward back propagation network (CFBPN) and radial basis function neural network (RBFNN) to verify the accuracy of the proposed method in harmonic source identification. Results proved that the identification of harmonic source at the point of common coupling using proposed method is more accurate with less computational time when compared to the other neural network structures.