Skip to content
Back to Projects

ReneWind Predictive Maintenance

Neural network classifier predicting wind turbine generator failures with ~97% accuracy and ~95% recall.

Shipping
ML Neural Networks · Python TensorFlow/Keras Scikit-learn Pandas NumPy
~97% Test Accuracy
~0.95 F1 Score (macro)
~95% Recall (macro)
7 Models Tested

Summary

Built a classification model to predict wind turbine generator failures using historical sensor data from 40 instruments per turbine. Tested 7 model configurations, with the best achieving ~97% test accuracy and ~95% recall on unseen data.

Approach

Trained neural networks with systematic architecture search (varying depth, optimizer, dropout, and class weighting) on imbalanced sensor data. Used walk-forward validation and focused on minimizing false negatives (missed failures) given the cost asymmetry: replacement >> repair >> inspection.

Key Result

Model 4 (Adam optimizer, 14→7 architecture) is recommended for deployment — it catches ~95% of impending failures while maintaining a ~3% false alarm rate.