how to load 3-phase voltage and current waveform data generated from simulink to workspace so as to use the data for training the neural network for detecting and classification of fault in a long transmission line

how to design neurel network and train them to detect ,classify and locate the ten types of fault on transmission line

Answer 1

[Greg Heath]

1. Extract I features from each of N examples and form the IxN dimensional input matrix.

2. For each example, assign a column from the 10-dimensional unit matrix eye(10) to indicate the fault type and form the OxN (O=10) dimensional output target matrix.

[ I N ] = size(input)

[ O N ] = size(target)

3. Consult the PATTERNNET help and documentation examples

 help patternnet

 doc patternnet

4. Search the NEWSGROUP and ANSWERS using the keyword patternnet.

5. Practice on one or more of the MATLAB classification/pattern-recognition data sets

 help nndatasets.

Thank you for formally accepting my answer

Greg