Article
Construction & Building Technology
Ke Yan
Summary: Data augmentation is essential for automated fault detection and diagnosis in chillers, and generative adversarial networks (GAN) can be used to generate synthetic faulty data samples. It has been shown that selecting high-quality synthetic faulty samples with GAN can improve FDD accuracy.
BUILDING AND ENVIRONMENT
(2021)
Article
Construction & Building Technology
Wanli Yao, Donghui Li, Long Gao
Summary: This paper proposes an effective fault detection and diagnosis method for centrifugal chillers by evaluating the performance of tree-based ensemble learning algorithms and multivariate control charts. The experimental results show that the proposed method can accurately identify various faults and achieve high detection rate and diagnosis accuracy, especially at low severity levels.
JOURNAL OF BUILDING ENGINEERING
(2022)
Article
Construction & Building Technology
Chengdong Li, Cunxiao Shen, Hanyuan Zhang, Hongchang Sun, Songping Meng
Summary: This paper proposes a novel feature-enhanced temporal convolutional network (FETCN) method for chiller fault diagnosis, which utilizes feature enhancement technique and statistical pooling to optimize fault diagnosis performance. Experimental results demonstrate that the method achieves a higher fault diagnosis rate at different fault severity levels and effectively addresses the dynamic coupling issue in chillers.
JOURNAL OF BUILDING ENGINEERING
(2021)
Article
Telecommunications
Ke Yan, Xiaokang Zhou
Summary: Computer-empowered fault detection plays a crucial role in HVAC systems, enhancing safety, saving energy, and protecting human health. Standard machine learning techniques have shown effectiveness in capturing chiller faults, but the utilization of deep learning approaches, such as convolutional neural networks, in chiller fault detection is still limited. This study proposes a CNN-based approach for chiller fault detection, which eliminates the need for feature selection and extraction processes and significantly improves classification accuracy.
DIGITAL COMMUNICATIONS AND NETWORKS
(2022)
Article
Telecommunications
Ke Yan, Xiaokang Zhou
Summary: Computer-empowered detection of possible faults for HVAC sub-systems, such as chillers, is a crucial application in Artificial Intelligence (AI) integrated Internet of Things (IoT). Standard machine learning techniques, such as Principal Component Analysis (PCA), Support Vector Machine (SVM) and tree-structure-based algorithms, are useful for capturing various chiller faults with high accuracy rates.
DIGITAL COMMUNICATIONS AND NETWORKS
(2022)
Article
Thermodynamics
Yudong Xia, Qiang Ding, Nijie Jing, Yijia Tang, Aipeng Jiang, Shu Jiangzhou
Summary: This paper presents an enhanced fault detection method using kernel density estimation and kernel entropy component analysis algorithms for effective detection of water chiller faults. By optimizing bandwidth and determining control limits for fault monitoring, the proposed method showed the best performance in experimental data validation, achieving a fault detection ratio of over 68% and an average fault detection accuracy of over 90%.
INTERNATIONAL JOURNAL OF REFRIGERATION
(2021)
Article
Automation & Control Systems
Wei-Ting Yang, Marco S. Reis, Valeria Borodin, Michel Juge, Agnes Roussy
Summary: Process monitoring is critical in manufacturing industries. This paper proposes an interpretable unsupervised machine learning model based on Bayesian Networks (BN) for fault detection and diagnosis. The model combines data-driven induction with domain knowledge and displays causal interactions in a graphical form. The proposed fault detection scheme consists of two levels of monitoring and uses local indices for fault diagnosis.
CONTROL ENGINEERING PRACTICE
(2022)
Article
Construction & Building Technology
Long Gao, Donghui Li, Ningyi Liang
Summary: This paper proposes a genetic algorithm (GA)-aided ensemble model without any characteristic assumption to detect and diagnose sensor fault in air-cooled chiller. The proposed method improves the detection and diagnosis performances by considering the Gaussian, non-Gaussian, non-linear, and dynamic characteristics simultaneously. Different statistical models are selected as sub-models in the ensemble framework to consider all the above characteristics. The proposed method delivers superior performance compared with other methods, as validated using data collected from a real chiller system.
BUILDING AND ENVIRONMENT
(2023)
Article
Engineering, Environmental
Nan Liu, Minggang Hu, Ji Wang, Yujia Ren, Wende Tian
Summary: Risks in chemical plants can be categorized as Black Swan events and Gray Rhino events. To address the challenge of root cause diagnosis in Gray Rhino events, a strong relevant mechanism Bayesian network (SRMBN) method is proposed for fault detection and diagnosis. The SRMBN is constructed through structure learning and parameter learning, and utilizes Bayesian interval estimated index and Bayesian contribution index for fault detection and diagnosis.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2022)
Article
Energy & Fuels
Bingxu Li, Fanyong Cheng, Xin Zhang, Can Cui, Wenjian Cai
Summary: A novel semi-supervised data-driven fault diagnosis method for chiller systems based on semi-generative adversarial networks is proposed to improve diagnostic performance using unlabeled data and reduce dependency on labeled data. Experimental results show that the proposed method significantly enhances diagnostic accuracy, even with limited labeled samples and vast unlabeled samples.
Article
Thermodynamics
Guannan Li, Liang Chen, Cheng Fan, Jiajia Gao, Chengliang Xu, Xi Fang
Summary: Fault diagnosis is crucial for maintaining the reliability and performance of chillers. In this study, a high-sensitivity gradient-based interpretation method was proposed to better interpret the CNN FD model for chiller faults. The method successfully localized fault-related feature variables and visualized diagnosis criteria, leading to improved diagnosis accuracy for early-stage faults.
APPLIED THERMAL ENGINEERING
(2024)
Article
Energy & Fuels
Hongwen Dou, Radu Zmeureanu
Summary: This paper presents the development and use of benchmarking grey-box models for the detection and diagnosis of multiple-dependent faults (MDFDD) of a water-cooled centrifugal chiller. The models are developed using data recorded by a Building Automation System (BAS) from a central cooling plant of an institutional building. The paper introduces a forward residual-based fault detection model and a rule-based backward approach for fault diagnosis. The proposed MDFDD model is tested with artificial faults inserted into the measurement data file, and the results showcase its potential for application.
Article
Construction & Building Technology
Viet Tra, Manar Amayri, Nizar Bouguila
Summary: This paper studies the fault detection and diagnosis for chillers in HVAC systems and proposes solutions for outliers detection and insufficiency of labeled data. The proposed approach, which includes the S-DAGMM algorithm and the use of DAGMM to pretrain DNN, achieves outstanding performance.
ENERGY AND BUILDINGS
(2022)
Article
Construction & Building Technology
Xu Zhu, Kang Chen, Burkay Anduv, Xinqiao Jin, Zhimin Du
Summary: This paper proposes a generic framework for transferring prior knowledge from a source chiller to build a diagnostic model for a new target chiller. The method involves standardizing heterogeneous data, applying domain adaption transfer learning to overcome domain shift, and utilizing domain adversarial neural network to generate the diagnostic model for the target chiller. Results show that the transferred diagnostic model yields decent performance, outperforming conventional machine learning models.
BUILDING AND ENVIRONMENT
(2021)
Article
Construction & Building Technology
Viet Tra, Manar Amayri, Nizar Bouguila
Summary: This paper emphasizes the importance of fault detection and diagnosis (FDD) for chillers in HVAC systems, and proposes a novel outlier detection (OD) method called NN-MPPCA. The experimental results demonstrate the superiority of NN-MPPCA in chiller FDD application.
BUILDING AND ENVIRONMENT
(2022)
Article
Energy & Fuels
Hong Tang, Shengwei Wang
Summary: This study develops a novel model-based predictive dispatch strategy for hybrid building energy systems to maximize economic benefits in electricity markets by considering the correlation between multiple flexibility resources and grid control signals, which can lower electricity costs in the electricity market.
Article
Energy & Fuels
Wenzhuo Li, Shengwei Wang
Summary: This paper proposes a fully distributed optimal control approach for HVAC systems to be implemented in IoT-enabled building automation networks. It utilizes the Incremental Cost Consensus (ICC) algorithm and the average consensus algorithm to optimize the individual rooms' outdoor air volume and estimate the outdoor air volume mismatch. Through tests and comparisons, it is found that the proposed approach with the fully connected topology outperforms existing hierarchical distributed approaches, demonstrating higher robustness, lower computation complexity, and higher optimization efficiency.
Article
Thermodynamics
Shaobo Sun, Shengwei Wang, Kui Shan
Summary: This study quantifies the measurement uncertainties of water flow meters in multiple water-cooled chiller systems using a Bayesian approach, proposing a quantification strategy that performs well in quantifying both systematic and random uncertainties with acceptable accuracy. The strategy can be used to optimize the control of chiller systems and improve their reliability.
APPLIED THERMAL ENGINEERING
(2022)
Article
Thermodynamics
Hangxin Li, Shengwei Wang
Summary: Constructing nearly and net zero-energy buildings presents the challenge of achieving energy goals post-occupancy, with traditional design methods showing high risk of failure. New ZEB design standards now focus on post-occupancy performance evaluation, posing additional challenges and calling for effective design methods.
BUILDING SIMULATION
(2022)
Article
Thermodynamics
Hangxin Li, Shengwei Wang
Summary: Model predictive control (MPC) method is superior in enhancing system performance, but is hindered by forecast uncertainties in operation. Different methods, such as shrinking horizon MPC (SHMPC) and stochastic MPC (SMPC), have been proposed to mitigate these uncertainties. However, there is limited knowledge about their year-round performance and relative performance, especially in the utilization of building flexibility-resources.
Article
Energy & Fuels
Shaobo Sun, Kui Shan, Shengwei Wang
Summary: This study proposed an online robust sequencing control strategy for chiller plants under low-quality and uncertain flow measurements, which effectively reduced the impacts of flow measurement uncertainties and improved the performance of chiller plants. The uncertainty processing model accurately quantified the measurement uncertainties of water flow rates, leading to significant reductions in root-mean-square error of cooling loads, total switching number of chillers, and cumulative unmet cooling load. The proposed control strategy showed the ability to tolerate flow measurement uncertainties.
Article
Construction & Building Technology
Hanbei Zhang, Fu Xiao, Chong Zhang, Rongling Li
Summary: This study develops a coordinated optimal load scheduling strategy for building cluster load management that considers dynamic electricity price and marginal emission factor simultaneously. The strategy effectively solves the conflicts of minimizing electricity cost, carbon emissions, and peak load while ensuring user satisfaction. The results show that the strategy can achieve a compromise between conflicting objectives in different correlation scenarios.
ENERGY AND BUILDINGS
(2023)
Article
Thermodynamics
Zhijie Chen, Fangzhou Guo, Fu Xiao, Xiaoyu Jin, Jian Shi, Wanji He
Summary: This study developed a data-driven benchmarking methodology to detect anomalous operations with degraded energy performance from a large number of bus ACs. By using a Long-Short-Term-Memory (LSTM) autoencoder-based similarity measurement method, similar operation data in other ACs can be identified for benchmarking. A LSTM network-based reference model is then developed to predict the power consumption of the target AC using these similar data. Statistical analysis-based trend and change detection algorithms are used to identify anomalies in power consumption ratio (PCR) over a few days. Two fault experiments were conducted to validate the methodology, showing its potential for health monitoring of bus ACs in a city fleet.
INTERNATIONAL JOURNAL OF REFRIGERATION
(2023)
Article
Engineering, Electrical & Electronic
Zhuang Zheng, Shengwei Wang, Wenzhuo Li, Xiaowei Luo
Summary: This paper proposes a novel voltage control strategy that regulates the on/off states of AC clusters to address voltage issues caused by high PV penetrations. The strategy includes temperature priority-based on/off control, real-time optimal demand response resources dispatch, distributed sensing of ACs, and flexibility capacity estimation. The strategy is validated to be effective and scalable, and is incorporated into a hierarchical control framework for smart grid voltage control.
IEEE TRANSACTIONS ON SMART GRID
(2023)
Article
Construction & Building Technology
Tianhang Zhang, Zilong Wang, Yanfu Zeng, Xiqiang Wu, Xinyan Huang, Fu Xiao
Summary: A novel framework of Artificial-Intelligence Digital Fire (AID-Fire) was proposed for real-time identification of building fire evolution, showing promising results in a full-scale fire test room.
JOURNAL OF BUILDING ENGINEERING
(2022)
Article
Construction & Building Technology
Wenxuan Zhao, Hangxin Li, Shengwei Wang
Summary: This study proposes a novel outdoor air ventilation strategy for high-tech cleanrooms, which enables maximum energy savings and efficiency. The strategy determines the optimal outdoor air volume by theoretically calculating the energy differential. Results show that the traditional fully coupled AHU system can achieve annual free cooling hours ranging from 662 to 2,537 in 31 major Chinese cities. Moreover, the proposed strategy achieves 8% energy savings in transition months and significant electricity and primary energy savings in a year.
BUILDING AND ENVIRONMENT
(2023)
Review
Construction & Building Technology
Xiaoyu Jin, Chong Zhang, Fu Xiao, Ao Li, Clayton Miller
Summary: Data related to building energy use is crucial for research and applications in building energy efficiency. However, most cities lack comprehensive and publicly accessible building energy use datasets, hindering urban building energy modeling, energy planning, performance benchmarking, and policymaking. This review paper provides insights based on a comprehensive analysis of worldwide open datasets and their applications, including detailed information on 33 building energy datasets and their subdomains, as well as proposed policy implications. The review also discusses privacy solutions and offers valuable conclusions to support city-level building energy data disclosure, modeling, and policymaking.
ENERGY AND BUILDINGS
(2023)
Article
Energy & Fuels
Ao Li, Chong Zhang, Fu Xiao, Cheng Fan, Yang Deng, Dan Wang
Summary: Data-driven models are widely used in smart building energy management. This paper investigates the performance of three conventional model update methods and five emerging continual learning methods using a 2-year dataset. The results show that continual learning methods are more effective in ensuring long-term accuracy while reducing computation time and data storage expenses.
Article
Thermodynamics
Yanxue Li, Zixuan Wang, Wenya Xu, Weijun Gao, Yang Xu, Fu Xiao
Summary: An efficient and flexible energy management strategy is crucial for energy conservation in the building sector. This study proposes a hybrid model-based reinforcement learning framework that uses short-term monitored data to optimize indoor thermal comfort and energy cost-saving performance. Simulation results demonstrate the efficiency and superiority of the proposed framework, with the D3QN agent achieving over 30% cost savings compared to measurement results.
Article
Computer Science, Artificial Intelligence
Xinqi Zhang, Jihao Shi, Xinyan Huang, Fu Xiao, Ming Yang, Jiawei Huang, Xiaokang Yin, Asif Sohail Usmani, Guoming Chen
Summary: This study proposes a deep probabilistic graph neural network that models the spatial dependency of sensors to improve leakage detection. The results demonstrate that the model achieves competitive detection accuracy and provides more comprehensive leakage detection information. Additionally, the model's posterior distribution enhances leakage localization accuracy.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Construction & Building Technology
Samiran Khorat, Debashish Das, Rupali Khatun, Sk Mohammad Aziz, Prashant Anand, Ansar Khan, Mattheos Santamouris, Dev Niyogi
Summary: Cool roofs can effectively mitigate heatwave-induced excess heat and enhance thermal comfort in urban areas. Implementing cool roofs can significantly improve urban meteorology and thermal comfort, reducing energy flux and heat stress.
ENERGY AND BUILDINGS
(2024)
Article
Construction & Building Technology
Qi Li, Jiayu Chen, Xiaowei Luo
Summary: This study focuses on the vertical wind conditions as a main external factor that limits the energy assessment of high-rise buildings in urban areas. Traditional tools for energy assessment of buildings use a universal vertical wind profile estimation, without taking into account the unique wind speed in each direction induced by the various shapes and configurations of buildings in cities. To address this limitation, the study developed an omnidirectional urban vertical wind speed estimation method using direction-dependent building morphologies and machine learning algorithms.
ENERGY AND BUILDINGS
(2024)
Article
Construction & Building Technology
Xiaojun Luo, Lamine Mahdjoubi
Summary: This paper presents an integrated blockchain and machine learning-based energy management framework for multiple forms of energy allocation and transmission among multiple domestic buildings. Machine learning is used to predict energy generation and consumption patterns, and the proposed framework establishes optimal and automated energy allocation through peer-to-peer energy transactions. The approach contributes to the reduction of greenhouse gas emissions and enhances environmental sustainability.
ENERGY AND BUILDINGS
(2024)
Article
Construction & Building Technology
Ying Yu, Yuanwei Xiao, Jinshuai Chou, Xingyu Wang, Liu Yang
Summary: This study proposes a dual-layer optimization design method to maximize the energy sharing potential, enhance collaborative benefits, and reduce the storage capacity of building clusters. Case studies show that the proposed design significantly improves the performance of building clusters, reduces energy storage capacity, and shortens the payback period.
ENERGY AND BUILDINGS
(2024)
Article
Construction & Building Technology
Felix Langner, Weimin Wang, Moritz Frahm, Veit Hagenmeyer
Summary: This paper compares two main approaches to consider uncertainties in model predictive control (MPC) for buildings: robust and stochastic MPC. The results show that compared to a deterministic MPC, the robust MPC increases the electricity cost while providing complete temperature constraint satisfaction, while the stochastic MPC slightly increases the electricity cost but fulfills the thermal comfort requirements.
ENERGY AND BUILDINGS
(2024)
Article
Construction & Building Technology
Somil Yadav, Caroline Hachem-Vermette
Summary: This study proposes a mathematical model to evaluate the performance of a Double Skin Facade (DSF) system and its impact on indoor conditions. The model considers various design parameters and analyzes their effects on the system's electrical output and room temperature.
ENERGY AND BUILDINGS
(2024)
Article
Construction & Building Technology
Ruijun Chen, Holly Samuelson, Yukai Zou, Xianghan Zheng, Yifan Cao
Summary: This research introduces an innovative resilient design framework that optimizes building performance by considering a holistic life cycle perspective and accounting for climate projection uncertainties. The study finds that future climate scenarios significantly impact building life cycle performance, with wall U-value, windows U-value, and wall density being major factors. By using ensemble learning and optimization algorithms, predictions for carbon emissions, cost, and indoor discomfort hours can be made, and the best resilient design scheme can be selected. Applying this framework leads to significant improvements in building life cycle performance.
ENERGY AND BUILDINGS
(2024)