Mechanical Engineering ETDs

Publication Date



The commercial building sector consumed about 20% of the total primary energy in the U.S. in 2008. A significant yet avoidable portion of the energy consumption is due to inefficient system operations. The inefficiencies can be attributed to degrading HVAC sub-systems, and undetected abnormal conditions. Recognition and remediation of these conditions through advanced data analytics can reduce energy consumption by 5% to 20%. This could save about $9 billion in utility costs in the U.S. alone. Modern buildings are constantly sending messages in the form of sensor data. However, this data is only as good as the system that collects it. Therefore, the present work explores fault detection and diagnostics (FDD) of an HVAC sub-system, in particular an air handling unit (AHU), through the evaluation of various methods. The detection methods include a controls alarm threshold, rule-based expressions, regression, one-class support vector machine (SVM), back-propagation, adaptive resonance theory (ART), and lateral priming adaptive resonance theory (LAPART). The diagnosis of AHU faults were performed using a multi-class SVM and LAPART algorithms. The results from the fault detection experiments were reviewed based on the two-class classification where the number of false positives and false negatives where compared. The diagnostic results were evaluated based on the comparison of precision and probability of detection values.


air handling unit, artificial neural networks, adaptive resonance theory, laterally primed adaptive resonance theory, support vector machines, fault detection, fault diagnostics, HVAC, machine learning

Degree Name

Mechanical Engineering

Level of Degree


Department Name

Mechanical Engineering

First Committee Member (Chair)

Caudell, Thomas

Second Committee Member

Tapia, Lydia

Third Committee Member

Sorrentino, Francesco


Yearout Mechanical Inc.

Document Type