Fourier Transform and Probabilistic Neural Network based Fault Detection in Distribution System Containing DGs

The protection relay may mal-operate with the penetration of distributed generation sources (DGs) in the distribution network. Classification of Fault is an important task for isolating the faulty section from distribution network on time. This paper represents a new combined approach of Discrete Fourier Transform (DFT) and Probabilistic Neural network (PNN) for the task of fault classification. Distribution network is modelled in PSCAD software in presence of multiple DGs. The DFT and PNN algorithm employed to classify the fault is coded in MATLAB software. Various fault cases are simulated to verify the validity of proposed scheme in different operating scenario. Moreover, various loading conditions are performed in presence of DGs and islanding conditions in presence of DGs. The result in term of fault classification precision proves that the proposed algorithm is competent to detect fault condition with better efficacy.