Comparative Analysis of Machine Learning Models for Predictive Maintenance in HVAC Constructions

Abstract

Predictive maintenance is vital for ensuring the longevity and optimal performance of building systems, such as Heating, Ventilation, and Air Conditioning (HVAC) units. By utilizing historical sensor data, machine learning techniques can forecast potential failures, enabling timely maintenance interventions. This study evaluates and compares the performance of three prominent machine learning models—Logistic Regression (LR), Random Forest (RF), and Gradient Boosting (GB)—in predicting HVAC unit failures. A simulated dataset encompassing features like temperature, humidity, pressure, usage hours, and maintenance logs was used for this comparison. Finally, Gradient Boosting demonstrated the highest performance, with an accuracy of 87%, precision of 0.86, recall of 0.88, F1-score of 0.87, and ROC-AUC of 0.92. These results underscore the superior predictive capabilities of Gradient Boosting and its potential in enhancing maintenance strategies for critical building systems. The results suggest that Gradient Boosting is the most effective model for enhancing predictive maintenance strategies in building systems, offering valuable insights for timely and efficient maintenance interventions.