Article
Engineering, Environmental
R. Rajesh Alias Harinarayan, S. Mercy Shalinie
Summary: This paper introduces the eXplainable Fault Detection, Diagnosis, and Correction (XFDDC) framework, which utilizes eXplainable Artificial Intelligence (XAI) techniques to explain the predictions of FDD models. The proposed framework is applied to the Tennessee Eastman Process (TEP) dataset and demonstrates the superior performance of the XGBoost model in fault detection.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2022)
Article
Green & Sustainable Science & Technology
Sara Ruiz-Moreno, Antonio J. Gallego, Adolfo J. Sanchez, Eduardo F. Camacho
Summary: Detecting and isolating faults in collector fields of solar thermal power plants is a crucial and challenging task that requires combining knowledge of systems engineering with machine learning techniques. Real irradiance profiles with different types of clouds were used for fault detection, and different machine learning techniques were compared, with the combination of neural networks being the only method that achieved high accuracy and F1-scores.
Article
Engineering, Mechanical
Lucas C. Brito, Gian Antonio Susto, Jorge N. Brito, Marcus A. Duarte
Summary: This paper introduces a new approach for fault detection and diagnosis in rotating machinery, which includes feature extraction, fault detection, and fault diagnosis. Fault detection is achieved through vibration feature extraction and anomaly detection algorithms, while fault diagnosis is performed using the SHAP technique for interpretation of black-box models.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Green & Sustainable Science & Technology
Shilin Sun, Tianyang Wang, Fulei Chu
Summary: A deep neural network method is proposed to address the data imbalance problem in wind turbine fault diagnosis. Convolutional and recurrent neural networks are used to extract spatial and temporal features from SCADA measurements. By employing a coarse learner and multiple fine learners, the reliability of fault diagnosis results is improved. A learner selection scheme is also designed to ensure computational efficiency. Experimental results demonstrate the effectiveness of the proposed method in improving accuracy and enhancing learning attention to all classes, making it a promising solution to wind turbine fault diagnosis.
Article
Automation & Control Systems
Dajian Huang, Wen-An Zhang, Steven X. X. Ding
Summary: Existing data-driven bearing fault diagnosis studies require complete fault samples, which is challenging to meet in the industry. This paper proposes a Cepstrum Scale-Distance based Framework (CSD-Framework) to address the issue of bearing fault diagnosis with incomplete training data. The framework includes three stages that transform vibration signals, adaptively adjust on multiple scales, and match distances using multiple metrics. The proposed method outperforms existing AI algorithms and Ceps-AI methods in terms of classification performance.
IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING
(2023)
Article
Thermodynamics
Cheng Fan, Yutian Lei, Yongjun Sun, Like Mo
Summary: Existing HVAC fault diagnosis methods relying on supervised learning are not feasible for individual buildings with limited labeled data. To address this, a novel self-supervised learning methodology based on transformers is proposed in this study, which extracts knowledge from unlabeled operational data for improved fault diagnosis performance. Data experiments using multiple HVAC datasets validate the efficacy of self-supervised learning, showing significant reduction in data labeling works and up to 8.44% improvement in fault diagnosis performance compared to conventional supervised learning. The research outcomes are valuable for developing high-performance data-driven solutions with limited labeled data in the building field.
Article
Construction & Building Technology
Dasheng Lee, Chih-Wei Lai, Kuo-Kai Liao, Jia-Wei Chang
Summary: The study proposed an innovative AI-assisted false alarm detection and diagnosis system, which can effectively reduce false alarms and improve the accuracy of fault detection. The system can meet high-reliability requirements and achieve significant maintenance cost savings.
JOURNAL OF BUILDING ENGINEERING
(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
Computer Science, Theory & Methods
Jose A. Ruiperez-Valiente, Daniel Jaramillo-Morillo, Srecko Joksimovic, Vitomir Kovanovic, Pedro J. Munoz-Merino, Dragan Gasevic
Summary: Collaboration is a key driver of learning and has been widely studied in various contexts, including MOOCs. A data-driven approach was used to detect and characterize collaboration in two large Coursera MOOCs, revealing different profiles of user accounts and types of collaborations. The findings suggest the presence of genuine collaborative associations as well as dishonest behaviors in online learning settings, calling for further research in this area.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2021)
Review
Education & Educational Research
Bahar Memarian, Tenzin Doleck
Summary: This article summarizes the pedagogical practices and tools used in data science education at the higher education level through a systematic literature review. The study finds that the content presented in data science education is diverse and difficult to compare. The study also examines the technological and pedagogical knowledge quality of the reviewed studies and lists the tools employed in each study.
EDUCATION AND INFORMATION TECHNOLOGIES
(2023)
Review
Health Care Sciences & Services
Grazia Dicuonzo, Graziana Galeone, Matilda Shini, Antonella Massari
Summary: This article explores the application of big data in healthcare and the integration of traditional data analytical tools and techniques for decision-making. The study results suggest that the acquisition, management, and analysis of a large volume of health data are crucial for effective and patient-centered care.
Article
Computer Science, Artificial Intelligence
Shilin Sun, Wenyang Hu, Yuekai Liu, Tianyang Wang, Fulei Chu
Summary: This paper proposes a novel method to address the problem of data imbalance in fault diagnosis of wind turbines. The method extracts spatial and temporal information from SCADA data and uses contrastive learning to obtain data representations correlated with the health conditions. The method achieves effective decision boundaries and impressive performance in recognizing faults in wind turbines.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Psychology, Multidisciplinary
Muhammad Usman Tariq, Muhammad Babar, Marc Poulin, Akmal Saeed Khattak, Mohammad Dahman Alshehri, Sarah Kaleem
Summary: The article discusses the evolution of intelligent big data analysis in the age of big data and artificial intelligence, focusing on the challenges of analyzing human behavior through social media data. It proposes an architecture to efficiently process massive social media datasets and demonstrates its effectiveness using data from Dailymotion.
FRONTIERS IN PSYCHOLOGY
(2021)
Review
Medicine, General & Internal
Arjan Sammani, Annette F. Baas, Folkert W. Asselbergs, Anneline S. J. M. te Riele
Summary: DCM is a leading cause of heart failure and LTVA, with great heterogeneity in phenotype and genotype making risk stratification challenging. Improved genetic testing has identified genotype-phenotype associations, allowing for better personalized risk assessments. Utilizing multivariable risk models, genetic risk scores, and advanced imaging techniques, as well as big data infrastructures and artificial intelligence, hold promise in enhancing predictive performance and prognosis of DCM.
JOURNAL OF CLINICAL MEDICINE
(2021)
Article
Engineering, Mechanical
Shihang Yu, Min Wang, Shanchen Pang, Limei Song, Xue Zhai, Yawu Zhao
Summary: This paper proposes a transferable decoupling multi-scale autoencoder (TDMSAE) to address the issue of poor adaptability of fault diagnosis models for mechanical equipment when the working environment changes. The proposed model utilizes a feature extraction module and a distribution alignment module to achieve higher accuracy in the target domain.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Engineering, Environmental
Nerea Uri-Carreno, Per H. Nielsen, Krist V. Gernaey, Qian Wang, Ulla Gro Nielsen, Marta Nierychlo, Susan H. Hansen, Lisette Thomsen, Xavier Flores-Alsina
Summary: Membrane-Aerated Biofilm Reactors (MABRs) are an effective alternative for wastewater treatment plants (WWTP) with enhanced nitrogen removal capacity and reduced energy consumption. This study demonstrates the impact of oxidation-reduction potential (ORP) on the performance of MABRs, showing that low ORP conditions can hinder nitrification. By increasing ORP and replacing/cleaning membranes, nitrification rates and the abundance of nitrifying organisms were significantly improved. The study also suggests that low ORP conditions promote the reduction of iron and sulfur compounds, leading to deposition in the biofilm and inhibiting nitrifiers growth.
CHEMICAL ENGINEERING JOURNAL
(2023)
Article
Computer Science, Interdisciplinary Applications
Simon B. Lindahl, Deenesh K. Babi, Krist V. Gernaey, Guerkan Sin
Summary: This paper proposes a methodology for setting up and solving integrated capacity and production planning problems. It utilizes mixed-integer linear programming problems to determine whether capacity levels should be changed. The methodology is applied to case studies from pharmaceutical manufacturing and demonstrates its ability to solve problems of industrial scale and relevance.
COMPUTERS & CHEMICAL ENGINEERING
(2023)
Article
Environmental Sciences
Xavier Flores-Alsina, Nerea Uri-Carreno, Per H. Nielsen, Krist V. Gernaey
Summary: Membrane Aerated Biofilm Reactors (MABR) are increasingly accepted in wastewater treatment processes. This study tests the prediction capabilities of an integrated modelling approach using full-scale data from a wastewater treatment plant. Results show a 10% mismatch between predictions and measurements, but the model provides valuable insights into biofilm structure and microbial activity.
SCIENCE OF THE TOTAL ENVIRONMENT
(2023)
Article
Engineering, Environmental
Morteza Zadkarami, Ali Akbar Safavi, Krist V. Gernaey, Pedram Ramin, Oscar A. Prado-Rubio
Summary: In recent years, water reclaim has become more relevant due to population increase and the number of industrial sites. Dynamic membrane filtration systems play an important role in water reuse. However, process monitoring of these systems is challenging due to intentional disturbances and input fluctuations, making fault detection and optimal operation difficult. This study presents a fault detection framework using classification methods applied to a pilot-scale dynamic ultrafiltration process. The results show that the Multilayer Perceptron Neural Network classifier has the highest detection accuracy and low false alarms. This framework is significant for digitalization and automated surveillance strategies for membrane system monitoring.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2023)
Article
Engineering, Environmental
Sebastian O. N. Topalian, Pedram Ramin, Kasper Kjellberg, Christian Kazadi Mbamba, Damien J. Batstone, Krist V. Gernaey, Xavier Flores-Alsina
Summary: In this study, a dynamic polymer dosing strategy is developed for an industrial centrifuge system that receives highly variable rates and quality of feed solids from a wastewater treatment plant. The strategy includes data extraction, data wrangling, model development, and model analysis, using partial least squares and random forest models to predict polymer dosages. The approach has the potential to save operators a significant amount of time and improve efficiency.
JOURNAL OF WATER PROCESS ENGINEERING
(2023)
Article
Education, Scientific Disciplines
Fiammetta Caccavale, Carina L. Gargalo, Krist V. Gernaey, Ulrich Kruhne
Summary: In times of educational disruption, significant advances have been made in adopting digitalization strategies. It is important for chemical engineers to be adequately educated to face the challenges of Industry 4.0 and acquire new skills. This study establishes a pedagogical framework to teach Python to chemical engineers, with courses covering topics such as chemical reaction engineering and machine learning. The students found the course to be useful and achieved practical problem-solving using Python.
EDUCATION FOR CHEMICAL ENGINEERS
(2023)
Article
Education, Scientific Disciplines
Isuru A. Udugama, Martin Atkins, Christoph Bayer, James Carson, Duygu Dikicioglu, Krist Gernaey, Jarka Glassey, Matthew Taylor, Brent R. Young
Summary: Educators in chemical engineering have a history of using digital tools for solving engineering problems, but it can be challenging to determine which tool is best for teaching specific concepts. A survey of department heads and committee members in IChemE institutions found that Microsoft Excel (VBA), commercial simulators, and scripting tools were ideal for teaching core subjects, while 3D models and Virtual/Augmented Reality were suited for teaching process design, safety, and sustainability. Non-technical factors such as simplicity, maintenance, and cost also play a role in tool adoption.
EDUCATION FOR CHEMICAL ENGINEERS
(2023)
Review
Environmental Sciences
Elham Ramin, Lourenco Faria, Carina L. Gargalo, Pedram Ramin, Xavier Flores-Alsina, Maj M. Andersen, Krist V. Gernaey
Summary: This study presents the first global review of water innovation practices in industrial symbiosis cases, proposing six categories of practices and highlighting regional variations and knowledge gaps, emphasizing the need for further research.
JOURNAL OF ENVIRONMENTAL MANAGEMENT
(2024)
Article
Computer Science, Interdisciplinary Applications
Carina L. Gargalo, Haoshui Yu, Nikolaus Vollmer, Ahmad Arabkoohsar, Krist Gernaey, Guerkan Sin
Summary: With the increasing concern for climate change, renewable and sustainable energy production has attracted considerable attention. Life cycle assessment (LCA) is an effective tool for comparing environmental impacts. However, there are differences in the choice of LCA methods, detail sharing, and sensitivity analysis among different studies, making it difficult to compare results.
COMPUTERS & CHEMICAL ENGINEERING
(2024)
Article
Engineering, Environmental
Xingxing Hu, Lingjie Liu, Yanmeng Bi, Lu Li, Chunsheng Qiu, Jingjie Yu, Shaopo Wang
Summary: In this study, the impact of exogenous folate on the start-up process of single-stage partial nitritation-anammox (SPNA) was evaluated using two lab-scale reactors. The results showed that folate addition can enhance nitrogen removal rate, extracellular polymeric substances production, hydrazine oxidase and dehydrogenase activity, as well as the relative abundance of Candidatus Brocadia.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2024)
Article
Engineering, Environmental
Shaocang He, Tingting Shen, Jing Sun, Haoqi Pan, Chenxu Sun, Tianpeng Li, Runyao Li, Enshan Zhang
Summary: A novel process of acid leaching neutralization was developed for the preparation of inorganic polymeric composite ferric aluminum silicate coagulant (CFAS) using solid waste coal gasification coarse slag (CGCS). The optimized preparation process was determined through single-factor experiment and the performance of CFAS was evaluated for domestic sewage treatment. The results showed that CFAS exhibited excellent coagulation ability and achieved significant removal efficiency for turbidity, ammonia nitrogen, and chemical oxygen demand.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2024)
Article
Engineering, Environmental
Kinga Szatmari, Sandor Nemeth, Alex Kummer
Summary: In this article, a resilience-based reinforcement learning approach is proposed to address the potential thermal runaway issue in batch reactors. By calculating the resilience metric for reactors and utilizing Deep Q-learning to decide when to intervene in the system, resilient-based mitigation systems can be effectively developed.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2024)
Article
Engineering, Environmental
Bingyou Jiang, Qi Yao, Mingqing Su, Jingjing Li, Kunlun Lu, Dawei Ding, Han Hong
Summary: This study investigates the inhibitory characteristics and mechanisms of ABC powder on coal powder explosion. The addition of ABC powder significantly decreases the maximum explosion pressure and can completely suppress coal dust explosions. The study also reveals the thermal decomposition characteristics and reaction kinetics of the mixed system.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2024)
Article
Engineering, Environmental
Kubilay Bayramoglu, Mustafa Nuran
Summary: This study examines the feasibility of using pyrolytic oil from waste tires as fuel in diesel engines, and evaluates its energy, exergy, and sustainability. The results indicate that pyrolytic oil has potential as a renewable fuel source with relatively high thermal efficiency.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2024)
Article
Engineering, Environmental
Xinru Duan, Yejia Lv, Jiaxing Hong, Jianzhong Wu, Jia Zhang, Yang Yue, Guangren Qian
Summary: This study successfully prepared a tube reactor with optimized catalyst formula, which showed good performance in removing dioxins and other pollutants in the experiments.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2024)
Article
Engineering, Environmental
Jiaoyang Du, Xueming Dang, Xiaorong Gan, Xin Cui, Huimin Zhao
Summary: In this study, a photocatalyst-enzyme hybrid system was constructed, which solved the issue of enzyme inactivation caused by high concentration of H2O2 through photocatalytic in-situ H2O2 production, and improved the stability and catalytic efficiency of the enzyme. The effectiveness of the system in treating phenolic EDCs was confirmed through experiments.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2024)
Article
Engineering, Environmental
Fu-Rong Xiu, Longsheng Zhan, Yingying Qi, Xinyue Lei, Jiali Wang, Haipeng Zhou, Wenting Shao
Summary: This study developed a synergetic and high-efficiency treatment of waste tantalum capacitors (WTCs) and polyvinyl chloride (PVC) using subcritical water process. The treatment significantly reduced the temperature and reaction time for metal tantalum recovery from WTCs, and improved the dechlorination efficiency of PVC. The optimized conditions resulted in 100% resin conversion efficiency of WTCs and 97.39% dechlorination efficiency of PVC. The interaction between decomposition products of WTCs and PVC produced a high level of benzoic acid.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2024)
Article
Engineering, Environmental
Bingda Li, Jiaming Song, Yuting Li, Chaoying Meng, Shuxian Wang, Linghao Zong, Honggang Ye, Yishuai Jing, Feng Teng, Peng Hu, Haibo Fan, Guangde Chen, Xin Zhao
Summary: CdPS3 nanosheets, especially those exfoliated by sodium cholate, have shown highly efficient photocatalytic degradation performance. The strong dark adsorption and dye-sensitized photocatalytic properties of CdPS3 nanosheets contribute to high degradation efficiencies of various pollutants.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2024)
Article
Engineering, Environmental
Dongxu Ouyang, Yimei Pang, Bo Liu, Zhirong Wang
Summary: This study investigates the thermal runaway features of lithium-ion cells under tunnel conditions, considering different states of charge and tunnel ceilings. The results show that the tunnel visibility is affected by the smoke generated during thermal runaway, and the shape of the tunnel ceiling influences the temperature rise differently.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2024)