A new Successive ANN for Fault Classification and Estimation of Combined Fault Resistance and Loading Conditions

The first part of this paper presents a new accurate approach for fault classification for high speed protective distance relaying using a new algorithm of ANN called Probabilistic Neural Network (PNN). This approach is based on phasor measurements of phase voltages and line currents as well as zero sequence currents. A comparison between the applied algorithm and other algorithms that have been used previously is presented. The second part of the paper proposes an algorithm for the estimation of both fault resistance and loading condition using a new back propagation technique called Scaled Conjugate Gradient (SCG) approach based on V-I phasor measurements and the type of fault that have been estimated in the first part. The required data measured and processed using MATLAB Simulink. The software tool chosen to handle the ANN is the (NNTOOL) of the MATLAB.