This repository contains the code used for the research study of Bearing Fault Detection Using Comparative Analysis of Random Forest, ANN, and Autoencoder Methods
As a machine is manufactured by an industry, with any product the first manufactured design is not completely efficient and can have several mishaps and faults. These can occur due to various reasons such as unfavorable weather conditions, extra exposure of the machine towards moisture. These can lead to motor malfunctions. These mishaps and faults can cause many unnecessary expenses and financial losses for the industry.
To prevent all these from happening a “Fault Detection for Predictive Maintenance in Industry 4.0” model has been designed as our project.
- Random Forest Classification
- Artificial Neural Network
- Autoencoders
- LSTM + Autoencoders
- K-means
- Isolation Forest
- One Class SVM
- Gaussian Distribution
- Principal Component Analysis