Enhancing Predictive Analysis of Wind Energy Potential Using Machine Learning
Students & Supervisors
Student Authors
Supervisors
Abstract
Research Area: This study investigates the use of machine learning (ML) models to predict wind energy generation based on data from wind turbine sensors. By leveraging advanced predictive modeling techniques, the research highlights how ML can enhance the accuracy of renewable energy forecasts, supporting more effective energy planning and management. Objectives Traditional statistical methods have been employed for forecasting wind energy potential [1] and often struggled to capture the complex, nonlinear dynamics involved in wind power generation. The objective of our study is to introduce a new dimension by utilizing machine learning models as forecasting tools and their capability to enhance prediction accuracy and reliability. Methodology The dataset [2] used in this study comprises wind turbine SCADA data containing features such as wind speed (m/s), wind direction (°), theoretical power output (kWh), and actual LV active power (kW). Selected relevant input features: Wind Speed, Wind Direction, and Theoretical Power Curve. Target variable: LV ActivePower (kW) and data was split into training (80%) and testing (20%) sets using train-test split. A total of ten Machine learniong models including Linear Regression, Ridge Regression, Lasso Regression, ElasticNet, Decision Tree Regressor, Random Forest Regressor, Gradient Boosting Regressor, AdaBoost Regressor, Support Vector Regressor (SVR) and K-Nearest Neighbors (KNN) Regressor were trained and tested. Results and Analysis The performance of ten machine learning models was evaluated using three key metrics: MAE, RMSE which is shown in Table 1, and R² Score which is shown in Figure 1. Among them, Gradient Boosting achieved the highest accuracy with the lowest RMSE (≈244.94) and the highest R² Score (≈0.9662), closely followed by Random Forest. In contrast, AdaBoost showed the weakest performance across all metrics. The table and graph clearly highlight the superiority of machine learning models in predicting wind energy potential."
Keywords
Publication Details
- Type of Publication: Conference
- Conference Name: IEEE Computer Society Bangladesh Chapter (CS BDC) Summer Symposium 2025
- Date of Conference: 18/07/2025 - 18/07/2025
- Venue: Hajee Mohammad Danesh Science & Technology University, Dinajpur
- Organizer: IEEE Compute society Bangladesh Chapter