Article
An Online Data-Driven Fault Diagnosis Method for Air
Handling Units by Rule and Convolutional Neural Networks
Huanyue Liao
1
, Wenjian Cai
2,
*, Fanyong Cheng
1
, Swapnil Dubey
3
and Pudupadi Balachander Rajesh
4
Citation: Liao, H.; Cai, W.; Cheng, F.;
Dubey, S.; Rajesh, P.B. An Online
Data-Driven Fault Diagnosis Method
for Air Handling Units by Rule and
Convolutional Neural Networks.
Sensors 2021, 21, 4358. https://
doi.org/10.3390/s21134358
Academic Editors:
Athanasios Rakitzis, Khanh T.
P. Nguyen and Kim Phuc Tran
Received: 5 May 2021
Accepted: 21 June 2021
Published: 25 June 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
1
SJ-NTU Corporate Lab, Nanyang Technological University, Singapore 637335, Singapore;
huanyue.liao@ntu.edu.sg (H.L.); Fanyong.cheng@nut.edu.sg (F.C.)
2
School of Electrical and Electronic Engineering, Nanyang Technological University,
Singapore 639798, Singapore
3
Energy Research Institute @ NTU, Singapore 637141, Singapore; sdubey@ntu.edu.sg
4
Surbana Jurong Consultants Pte Ltd., Singapore 150168, Singapore; rajesh.balachander@surbanajurong.com
* Correspondence: ewjcai@ntu.edu.sg; Tel.: +65-6790-6862
Abstract:
The stable operation of air handling units (AHU) is critical to ensure high efficiency
and to extend the lifetime of the heating, ventilation, and air conditioning (HVAC) systems of
buildings. In this paper, an online data-driven diagnosis method for AHU in an HVAC system is
proposed and elaborated. The rule-based method can roughly detect the sensor condition by setting
threshold values according to prior experience. Then, an efficient feature selection method using 1D
convolutional neural networks (CNNs) is proposed for fault diagnosis of AHU in HVAC systems
according to the system’s historical data obtained from the building management system. The new
framework combines the rule-based method and CNNs-based method (RACNN) for sensor fault
and complicated fault. The fault type of AHU can be accurately identified via the offline test results
with an accuracy of 99.15% and fast online detection within 2 min. In the lab, the proposed RACNN
method was validated on a real AHU system. The experimental results show that the proposed
RACNN improves the performance of fault diagnosis.
Keywords: convolutional neural network; HVAC system air handling unit; fault diagnosis
1. Introduction
Heating, ventilation, and air conditioning (HVAC) systems are the major energy
consumers in the building, which contribute to over 40% of the energy consumption [
1
–
4
].
The reliability and availability of the HVAC system needs to be further improved in order
to increase energy efficiency with proper operation and maintenance [
5
–
7
]. The large-scale
HVAC systems have complex structures and usually work under unpredictable weather
conditions and different indoor user settings [
2
]. Therefore, the actual operating state
of the HVAC system is rather complex, and the features of the state are variable with
strong nonlinearity and coupling characteristics [
8
]. Additionally, the HVAC systems
include large sums of equipment such as dampers, sensors, controlled actuators, and
so on. If maintenance is not implemented in a timely manner if a minor fault happens,
this will reduce the efficiency, accelerate the equipment deterioration, and even damage
other components [
9
]. Studies have shown that the highest failure of HVAC systems can
lead to an increase in energy consumption of 30% [
10
]. Hence, it is essential to apply the
fault diagnosis or the HVAC system to improve its reliability and performance. The air
handling unit (AHU) is the main functioning component of the entire HVAC system, which
is responsible for controlling the temperature, humidity, flow rate, and so on. The reliability
of AHU is critical for the operation and maintenance of the HVAC system [
11
]. Accurate
and fast fault detection and diagnosis are highly desirable for AHU in HVAC systems.
Fault detection and diagnosis have always been critical problems in the efficient and
reliable operation of HVAC and energy optimization. In the past 20 years, research on
Sensors 2021, 21, 4358. https://doi.org/10.3390/s21134358 https://www.mdpi.com/journal/sensors
