Data-driven fault detection and diagnosis for packaged rooftop units using statistical machine learning classification methods

Ebrahimifakhar, Amir; Kabirikopaei, Adel; Yuill, David P.

doi:10.1016/j.enbuild.2020.110318

Cited by 49 publications

(18 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The comparison between real-time data stream and learned historical patterns yields accurate operation diagnosis for a few fault types in a real building. Later in the past decade, various aspects of ML methods were intensely explored to characterize occurrences of faults, detect abnormal operating conditions, and classify fault types, such as adaptive thresholds, using t-statistic approach, fuzzy logic, ANN, Gaussian process regression, support vector machine, , gradient boosting regression, and generative adversarial network . However, these methods rarely captured temporal dependencies and dynamics of faults.…”

Section: Performance Prediction and Design Optimization Of Thermal En...mentioning

confidence: 99%

Machine Learning for Harnessing Thermal Energy: From Materials Discovery to System Optimization

Dai

2022

ACS Energy Lett.

View full text Add to dashboard Cite

Recent advances in machine learning (ML) have impacted research communities based on statistical perspectives and uncovered invisibles from conventional standpoints. Though the field is still in the early stage, this progress has driven the thermal science and engineering communities to apply such cutting-edge toolsets for analyzing complex data, unraveling abstruse patterns, and discovering non-intuitive principles. In this work, we present a holistic overview of the applications and future opportunities of ML methods on crucial topics in thermal energy research, from bottom-up materials discovery to top-down system design across atomistic levels to multi-scales. In particular, we focus on a spectrum of impressive ML endeavors investigating the state-of-the-art thermal transport modeling (density functional theory, molecular dynamics, and Boltzmann transport equation), different families of materials (semiconductors, polymers, alloys, and composites), assorted aspects of thermal properties (conductivity, emissivity, stability, and thermoelectricity), and engineering prediction and optimization (devices and systems). We discuss the promises and challenges of current ML approaches and provide perspectives for future directions and new algorithms that could make further impacts on thermal energy research.

show abstract

Section: Performance Prediction and Design Optimization Of Thermal En...mentioning

confidence: 99%

Machine Learning for Harnessing Thermal Energy: From Materials Discovery to System Optimization

Dai

2022

ACS Energy Lett.

View full text Add to dashboard Cite

show abstract

“…The DD methods leverage historical labeled data and powerful models from AI techniques to perform multi-class regression or classification, which are quite important for detecting faults [21]. In the literature, FDD methods using AI models such as random forest [22], collaborative filtering [23], extreme gradient boosting [24], and such techniques have been proposing (see, [25] and references therein). Despite these advances, detecting incipient faults addressing the fundamental challenges of accuracy at low irradiance conditions, lack of explainability on decisions to field engineers/technicians, and distinguishing false alarms from incipient faults is rather unexplored in the literature to our best knowledge.…”

Section: B Literature Reviewmentioning

confidence: 99%

Explainable Incipient Fault Detection Systems for Photovoltaic Panels

Sairam,

Srinivasan,

Marafioti

et al. 2020

Preprint

View full text Add to dashboard Cite

This paper presents an eXplainable Fault Detection and Diagnosis System (XFDDS) for incipient faults in PV panels. The XFDDS is a hybrid approach that combines the modelbased and data-driven framework. Model-based FDD for PV panels lacks high fidelity models at low irradiance conditions for detecting incipient faults. To overcome this, a novel irradiance based three diode model (IB3DM) is proposed. It is a nine parameter model that provides higher accuracy even at low irradiance conditions, an important aspect for detecting incipient faults from noise. To exploit PV data, an extreme gradient boosting (XGBoost) is used due to its ability to detecting incipient faults. Lack of explainability, feature variability for sample instances, and false alarms are challenges with data-driven FDD methods. These shortcomings are overcome by hybridization of XGBoost and IB3DM, and using eXplainable Artificial Intelligence (XAI) techniques. To combine the XGBoost and IB3DM, a faultsignature metric is proposed that helps reducing false alarms and also trigger explanation on detecting incipient faults. To provide explainability, an eXplainable Artificial Intelligence (XAI) application is developed. It uses the local interpretable model-agnostic explanations (LIME) framework and provides explanations on classifier outputs for data instance. These explanations help field engineers/technicians for performing troubleshooting and maintenance operations. The proposed XFDDS is illustrated using experiments on different PV technologies and our results demonstrate the perceived benefits.

show abstract

“…Faults in RTUs are categorized into two main types: (a) hard faults, also called hard failures, which cause the RTU to stop functioning and (b) soft faults, which can decrease the performance of the RTU until a hard failure occurs. One advantage of a fault detection and diagnosis (FDD) system is that it can detect and diagnose soft faults before hard failures occur [2] . A common approach to detecting faults in HVAC systems is to collect real-time operating characteristic data using a sensor network and analyze the collected data to determine if faults are present and the type of fault.…”

Section: Introductionmentioning

confidence: 99%

“…FDD tools and methods have been developed extensively using laboratory data [2][3][4][5][6] . As an example, in their laboratory study, Braun and Yuill [3] developed a methodology to assess the FDD protocols for air conditioning devices.…”

Section: Introductionmentioning

confidence: 99%

“…The accuracy of their method was 80-89% with a training dataset of 8000 normal samples and 30 faulty samples for each fault. New samples were artificially generated to balance the minority classes or faults using the Synthetic Minority Over-sampling Technique (SMOTE) [2,12] . Imbalanced data can significantly influence the fault classification accuracy.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Semi-Supervised Machine Learning for Fault Detection and Diagnosis of a Rooftop Unit

Albayati

Faraj

Thompson

et al. 2023

Big Data Min. Anal.

View full text Add to dashboard Cite

Most heating, ventilation, and air-conditioning (HVAC) systems operate with one or more faults that result in increased energy consumption and that could lead to system failure over time. Today, most building owners are performing reactive maintenance only and may be less concerned or less able to assess the health of the system until catastrophic failure occurs. This is mainly because the building owners do not previously have good tools to detect and diagnose these faults, determine their impact, and act on findings. Commercially available fault detection and diagnostics (FDD) tools have been developed to address this issue and have the potential to reduce equipment downtime, energy costs, maintenance costs, and improve occupant comfort and system reliability. However, many of these tools require an in-depth knowledge of system behavior and thermodynamic principles to interpret the results. In this paper, supervised and semi-supervised machine learning (ML) approaches are applied to datasets collected from an operating system in the field to develop new FDD methods and to help building owners see the value proposition of performing proactive maintenance. The study data was collected from one packaged rooftop unit (RTU) HVAC system running under normal operating conditions at an industrial facility in Connecticut. This paper compares three different approaches for fault classification for a real-time operating RTU using semi-supervised learning, achieving accuracies as high as 95.7% using few-shot learning.

show abstract

Data-driven fault detection and diagnosis for packaged rooftop units using statistical machine learning classification methods

Cited by 49 publications

References 25 publications

Machine Learning for Harnessing Thermal Energy: From Materials Discovery to System Optimization

Machine Learning for Harnessing Thermal Energy: From Materials Discovery to System Optimization

Explainable Incipient Fault Detection Systems for Photovoltaic Panels

Semi-Supervised Machine Learning for Fault Detection and Diagnosis of a Rooftop Unit

Contact Info

Product

Resources

About