Predicting Wind Turbine Power Production Using SCADA Data

This project aimed to develop a machine learning model that accurately predicts wind turbine power production using SCADA data. The project utilized various tools and techniques, including Python, exploratory data analysis (EDA), data wrangling, and the Random Forest algorithm.

The project analyzed SCADA data, including wind speed, wind direction, theoretical power values, and actual power generated by the turbine. Python was used for exploratory data analysis and data wrangling to prepare the data for modeling. Various types of regressor algorithms (using Scikit-Learn) were then trained on the prepared data to predict the power production of wind turbines.

The best resulting model was deployed on a Streamlit app, enabling users to input wind speed and wind direction values and receive a predicted power production value. Overall, this project demonstrated how machine learning can optimize energy production in the renewable energy sector. The use of Python, EDA, data wrangling, and the Random Forest algorithm highlighted how these tools could be utilized to develop precise predictive models.