Oil-injection-Screw-Compressor-Fault-Classification

1. This is a Machine Learning Program to classify Screw-Compressor Faults using Short Time Fourier Transform Spectrograms
2. Vibration signals are converted to spectrograms.
3. A Deep Learning Model is trained to predict fault states using spectrograms.

Algorithm

Fault Description

There are total of 5 states for the model to predict.

1. Normal Condition
2. Unbalance Condition
3. Belt Looseness Condition
4. Belt Looseness High Condition
5. Bearing Fault Condition

References

1. Dataset : PHM 2021 Asia Pacific Data Challenge