Implementing the approach proposed by the research article Deep Learning Enabled Fault Diagnosis Using Time-Frequency Image Analysis of Rolling Element Bearings in a tool for predictive maintenance using deep neural network.
We use Python with a standard virtualenv
setup.
The file requirements.txt
contains all the dependencies needed to run the code.
For the Jupyter Notebooks we recommend running them on Google Colaboratory, utilizing GPU acceleration. Make sure the Notebook runtime is GPU accelerated by clicking "Runtime" > "Change runtime type" from the top level menu.
Alternatively one can use any Jupyter Environment. In this case the first cells in the Notebooks must be ignored/adjusted, as they are specific to the Colab environment.
Just follow the Jupyter Notebooks, starting with colab/deeppredict_0_download_raw_data.ipynb
.
Make sure you have a writable DEEPPREDICT_HOME
folder with the same contents and structure as this repository.