how to prepare numeric data for classification of faults to import in deep network designer app in matlab

how to prepare numeric data for classification of faults in MMC to import in deep network designer app in matlab

Answer 1

[dpb]

That's a big Q? with no background...I'd suggest this could be a good place to use <MATLAB Copilot> as well as look at the examples provided for something that seems at least somewhat simimlar to whatever it is you have in mind.

Answer 2

[Arkadiy Turevskiy]

Hi. Does MMC stanf for modular multilevel converter?

For fault classification people often start with extracting features from time-series data. Once you have features you can use them for fault classification.

For feature extraction I recommend checking out Diagnostic Feature Designer app in Predictive Maintenance Toolbox, please take a look at the examples here: https://www.mathworks.com/help/predmaint/examples.html?s_tid=CRUX_topnav&category=rotating-machinery

This example in particular shows how to use extracted features for classification with a deep neural network (you can specify network architecture using Deep Network Designer app or programmatically as the example does):

https://www.mathworks.com/help/predmaint/ug/rolling-element-bearing-fault-diagnosis-using-deep-learning.html

HTH.

Arkadiy