Supervised machine learning methods for condition monitoring and fault detection of wind turbines

This is my code repository for my Engineering Thesis. The goal of the thesis was to use three Supervised Machine Learning algorithms, being Support-vector machine, k-Nearest neighbors and Gaussian process regression to detect faults and predict abnormalities in behaviour of a wind turbine.

Dataset

Data was acquired from a wind farm based in La Haute-Borne, France.
To acquire data used for this project, download .CSV file from this website:

https://opendata-renewables.engie.com/explore/index

Abbreviations

Gb1t_SVM - Gearbox bearing temperature Support vector machine
Db1t_SVM - Generator bearing temperature Support vector machine
Gb1t_KNN - Gearbox bearing temperature K-Nearest neighbors
Db1t_KNN - Generator bearing temperature K-Nearest neighbors
Gb1t_GPR - Gearbox bearing temperature Gaussian process regression
Db1t_GPR - Generator bearing temperature Gaussian process regression