Article
Materials Science, Multidisciplinary
Stylianos Tsopanidis, Shmuel Osovski
Summary: Modern computer vision and machine learning techniques have the potential to automate much of the failure analysis process in Fractography and remove human-induced ambiguity or bias. Deep learning methods, efficient in establishing complex interconnections between input data, may reveal new correlations and information encoded onto the complex geometries of fracture surfaces. The use of unsupervised learning to classify fracture surfaces based on tungsten percentage has shown promising results, with plasticity on the fracture surface serving as a measure for classification.
MATERIALS CHARACTERIZATION
(2021)
Article
Mathematics
Mariana-Ioana Maier, Gabriela Czibula, Zsuzsanna-Edit Onet-Marian
Summary: By comparing traditional and synchronous online learning methods, this study found that autoencoders can detect hidden patterns in academic data sets unsupervised and are valuable for predicting students' performance. The results showed that while traditional evaluations are slightly more accurate than online evaluations, there was no statistically significant difference between the two types of assessments.
Article
Environmental Studies
James Ming Chen, Mobeen Ur Rehman, Xuan Vinh Vo
Summary: This article applies unsupervised machine learning to visualize and interpret logarithmic returns and conditional volatility in commodity markets. Results show that returns-based clustering conforms closely to traditional boundaries between different types of commodities, while volatility-based clustering successfully identifies extreme market distress periods.
Article
Engineering, Civil
Muhammad Kamran, Niaz Muhammad Shahani, Danial Jahed Armaghani
Summary: Coal pillar assessment is crucial for underground engineering structures due to the potential disasters caused by pillar failure. Traditional forecasting techniques are insufficient in generating accurate outcomes due to the non-linear correlation between pillar failure and its influential attributes. This paper proposes a new approach using combined unsupervised-supervised learning to forecast underground coal pillar stability. By building a database of authentic engineering structures and employing advanced feature depletion and clustering methods, the proposed model can accurately predict the class of pillar failure in various underground rock engineering projects.
GEOMECHANICS AND ENGINEERING
(2022)
Article
Computer Science, Artificial Intelligence
Mudeet Jain, Mehul Mahrishi, Girish Sharma, Samira Hosseini
Summary: India saw a 23% increase in podcast listening during the Covid-19 pandemic. People turned to their favorite old audio podcasts due to the pandemic and screen fatigue. The classification of podcast genres helps listeners create playlists and enables podcast streaming services to recommend relevant content based on the genres users enjoy.
Article
Automation & Control Systems
Amirhossein Rahbari, Marc Rebillat, Nazih Mechbal, Stephane Canu
Summary: Structural Health Monitoring (SHM) is a challenging task in industrial fields like aeronautics due to rare and costly damaged data. Unsupervised dimensionality reduction techniques are appealing but unable to cluster unknown samples. By associating projection bases with Deep Neural Networks (DNNs, the method can effectively cluster any incoming unknown samples.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2021)
Article
Chemistry, Multidisciplinary
Dominic Owusu-Ansah, Joaquim Tinoco, Faramarzi Lohrasb, Francisco Martins, Jose Matos
Summary: This paper presents an alternative approach to predict rockburst using Machine Learning (ML) algorithms. The study used the Decision Tree (DT) algorithm and implemented two approaches: DT-RT and Unique-DT. A dataset containing 210 records was collected. Training and testing were performed on this dataset with 5 input variables. Other ML algorithms, such as RF, SVM, ANN, KNN, and AdaboostM1, were implemented for comparison. The evaluation metrics and relative importance were utilized to examine the characteristics of the DT methods. The Unique-DT model showed promising results with an average F1 score of 0.65, making it recommended for predicting rockburst conditions due to its ease, efficiency, and accuracy.
APPLIED SCIENCES-BASEL
(2023)
Article
Green & Sustainable Science & Technology
Guangtuo Bao, Kepeng Hou, Huafen Sun
Summary: A rock burst intensity-grade prediction model based on the comprehensive weighting of prediction indicators and Bayesian optimization algorithm-improved-support vector machine (BOA-SVM) is proposed to accurately judge the tendency of rock burst disaster. The model shows higher accuracy and better effect than ordinary models, providing a new way of thinking for rock burst intensity-grade prediction.
Article
Computer Science, Artificial Intelligence
Zhiwang Zhang, Hongliang Sun, Shuqing Li, Jing He, Jie Cao, Guanghai Cui, Gang Wang
Summary: This paper proposes a novel two-stage sparse multi-kernel optimization classifier (TSMOC) method to address the issue of inconsistent classification caused by redundancy and unrelated attributes. By introducing single or multi-kernel functions into classifier models, non-linearly separable problems are solved, but predictive interpretability is reduced. Experimental results show that TSMOC outperforms seven other classifiers on thirteen real-world datasets.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Social Sciences, Interdisciplinary
Adriano de Oliveira Andrade, Leonardo Garcia Marques, Osvaldo Resende, Geraldo Andrade de Oliveira, Leandro Rodrigues da Silva Souza, Adriano Alves Pereira
Summary: This study describes the conception and evaluation of a system for defining and predicting project profiles. By analyzing data from government-funded SICONV projects in Brazil, data clustering was achieved and ten project profiles were defined. Among multiple prediction models, the k-Nearest-Neighbor model performed the best with high accuracy.
Article
Computer Science, Information Systems
Jianhui Lv, Qing Li, Yong Jiang
Summary: This paper investigates the classical cache allocation problem of distributing cache capacity to Content Routers under a constrained and fixed total cache budget by considering network topology information and traffic characteristics. It uses t-distributed Stochastic Neighbor Embedding (t-SNE) to reduce data dimensions, divides Content Store into collaborative and non-collaborative regions, and achieves better efficiency in cache allocation strategy.
Article
Metallurgy & Metallurgical Engineering
Kasimcan Koruk, Julian M. Ortiz
Summary: Building a geological model is crucial for resource estimation, and utilizing geochemical data and machine learning techniques can lead to more accurate models.
MINING METALLURGY & EXPLORATION
(2023)
Article
Engineering, Electrical & Electronic
Jiajing Zhou, Zhao An, Zhile Yang, Yanhui Zhang, Huanlin Chen, Weihua Chen, Yalin Luo, Yuanjun Guo
Summary: This paper proposes PT-Informer, a deep learning framework for fault prediction and detection in nuclear power plants. Unlike traditional approaches, PT-Informer extracts fault features from raw vibration signals and achieves ultra-real-time fault prediction. Experimental results demonstrate that PT-Informer outperforms traditional models in terms of prediction accuracy and fault classification.
Article
Computer Science, Artificial Intelligence
Chuanbing Wan, Fusheng Jin, Zhuang Qiao, Weiwei Zhang, Ye Yuan
Summary: This paper proposes a novel unsupervised active learning method that models the nonlinearity of data using a deep neural network and considers essential criteria such as representativeness, informativeness, and diversity. Experimental results demonstrate the effectiveness of the method.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Mathematics
Barkat Ullah, Muhammad Kamran, Yichao Rui
Summary: Accurate prediction of short-term rockburst is crucial for improving worker safety in mining and geotechnical projects. This study utilized t-distributed stochastic neighbor embedding (t-SNE), K-means clustering, and extreme gradient boosting (XGBoost) to predict the risk of short-term rockburst. The proposed model achieved high accuracy in predicting rockburst levels, providing a valuable benchmark for future predictions.
Article
Ophthalmology
K. T. Boden, A. Rickmann, F. N. Fries, K. Xanthopoulou, D. Alnaggar, K. Januschowski, B. Seitz, B. Kaesmann-Kellner, J. Schrecker
Article
Engineering, Geological
Mohammadali Sepehri, Derek B. Apel, Samer Adeeb, Paul Leveille, Robert A. Hall
ENGINEERING GEOLOGY
(2020)
Article
Engineering, Geological
Yuanyuan Pu, Derek B. Apel, Robert Hall
ENGINEERING GEOLOGY
(2020)
Article
Geosciences, Multidisciplinary
Yuanyuan Pu, Derek B. Apel, Stanislaw Prusek, Andrzej Walentek, Tomasz Cichy
Summary: By using decision tree regressor and neural network models, the relationship between ground stress field and its impact factor can be modeled with limited field measurement data. This allows for the back-analysis of initial stress and the prediction of geological hazards, supporting the construction of underground engineering projects.
Review
Engineering, Geological
Jun Wang, Derek B. Apel, Yuanyuan Pu, Robert Hall, Chong Wei, Mohammadali Sepehri
Summary: As excavation depth increases, rockburst has become a serious geological hazard, prompting researchers worldwide to investigate different methods to address this problem. Numerical modeling, aided by advancements in IT and computer technology, has made significant progress in studying rockbursts.
JOURNAL OF ROCK MECHANICS AND GEOTECHNICAL ENGINEERING
(2021)
Article
Computer Science, Artificial Intelligence
Yuanyuan Pu, Jie Chen, Derek B. Apel
Summary: This study aims to predict the time to the next failure in laboratory fault failure experiments using acoustic emission (AE) signals and a customized deep learning network. The network achieved satisfactory results and showed promise for earthquake prediction. Further studies are needed for industrial applications.
NEURAL COMPUTING & APPLICATIONS
(2021)
Article
Metallurgy & Metallurgical Engineering
Jun Wang, Derek B. Apel, Artur Dyczko, Andrzej Walentek, Stanislaw Prusek, Huawei Xu, Chong Wei
Summary: Excavation-induced stresses from upper coal seams play a key role in rockbursts during close-distance coal seam mining, with multiple stresses contributing to stress concentration and energy accumulation. External causes like side abutment stress from mining in the upper coal seam and deviatoric stress from complex excavating situations also influence rockburst occurrence. Calibration of indicators is essential for predicting rockburst potential, and a methodology combining numerical modeling, laboratory tests, and field feedback can help estimate and mitigate rockbursts in similar underground mining conditions.
MINING METALLURGY & EXPLORATION
(2021)
Article
Engineering, Civil
Huawei Xu, Derek B. Apel
Summary: This study compares the sidewall swellings, floor heaves, and roof subsidence in crossing cuts of stopes under different last mined stope location scenarios using the Finite Element Method. The results show that choosing a rational location for the last mined stope can effectively reduce the instability risks caused by sill pillar recovery.
INTERNATIONAL JOURNAL OF GEOMATE
(2021)
Article
Energy & Fuels
Lukasz Wojtecki, Sebastian Iwaszenko, Derek B. Apel, Tomasz Cichy
Summary: This article explores the use of machine learning algorithms to predict rockburst risk in underground coal mines. By utilizing various algorithms, rockburst risk prediction models were proposed. Neural network and decision tree models were found to be most effective in this aspect.
Article
Engineering, Geological
Lukasz Wojtecki, Sebastian Iwaszenko, Derek B. Apel, Miroslawa Bukowska, Janusz Makowka
Summary: This study uses machine learning algorithms to assess the rockburst hazard in a hard coal mine. The decision tree and neural network models are proven to be effective in distinguishing rockbursts from tremors with high accuracy.
JOURNAL OF ROCK MECHANICS AND GEOTECHNICAL ENGINEERING
(2022)
Article
Metallurgy & Metallurgical Engineering
Chong Wei, Derek Apel, Tatyana Katsaga
Summary: This study presents an effective determination process of interface microstructural parameters and modeling mechanical interactions between different geomaterials with frictional-bonded interfaces using the coupled finite-difference and discrete-element method. Direct shear tests and parametric studies were conducted to validate the feasibility and rationality of the coupled FD-DE method in simulating specimens with frictional-bonded interfaces or joints.
MINING METALLURGY & EXPLORATION
(2022)
Article
Energy & Fuels
Huawei Xu, Derek B. Apel, Jun Wang, Chong Wei, Krzysztof Skrzypkowski
Summary: In this study, the feasibility of recovering sill pillars in a hard rock mine was investigated. Three recovery schemes were proposed and evaluated using a full-sized three-dimensional analysis model and the finite element method. The results showed that all three schemes were feasible and safe, with scheme SBS being the optimum one.
Article
Energy & Fuels
Krzysztof Skrzypkowski, Krzysztof Zagorski, Anna Zagorska, Derek B. Apel, Jun Wang, Huawei Xu, Lijie Guo
Summary: The article presents a method for selecting arch yielding support for preparatory workings driven in a hard coal seam. The impact of fault-induced discontinuous deformation on excavation protection schemes is investigated. Analytical calculations and numerical simulations are conducted to determine the displacements and reinforcement options in the fault zone. The research findings demonstrate the feasibility of pre-selecting reinforcement strategies to minimize movement dimensions during excavation.
Article
Energy & Fuels
Jun Wang, Derek B. Apel, Huawei Xu, Chong Wei, Krzysztof Skrzypkowski
Summary: In this paper, a 2D distinct element method (DEM) model is used to evaluate the effects of different types of rockbolts on controlling self-initiated strainbursts. The results show that D-bolts and Roofex are more effective in controlling strainbursts compared to resin-grouted rebar.
Review
Energy & Fuels
Prerita Odeyar, Derek B. Apel, Robert Hall, Brett Zon, Krzysztof Skrzypkowski
Summary: This paper provides a comprehensive review of different statistical techniques used for reliability and fault prediction, discussing their advantages, limitations, and comparing them to traditional methods. Researchers are working on new methods to analyze faults and improve reliability.
Article
Construction & Building Technology
Zhi Ding, Xiao Zhang, Shao-Heng He, Yong-Jie Qi, Cun-Gang Lin
Summary: This study investigates the longitudinal behavior of a shield tunnel by designing and constructing a reduced-size indoor model. The results show that the longitudinal settlement of the tunnel follows a normal distribution, with the maximum settlement occurring at the central ring and increasing linearly with the applied load. Stress concentration typically occurs on the side of the tunnel waist under surcharge, resulting in transverse elliptical deformation of the entire structure.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Lucia Lopez-de-Abajo, Marcos G. Alberti, Jaime C. Galvez
Summary: Assessing and predicting concrete damage is crucial for infrastructure management. This study quantifies gas concentrations in urban tunnels to achieve this goal.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Chao He, Yinghao Cai, Chenqiang Pu, Shunhua Zhou, Honggui Di, Xiaohui Zhang
Summary: This paper investigates the impact of river channel excavation on adjacent metro tunnels and proposes protective measures based on an engineering project in Fuzhou, China. A three-dimensional finite element model is developed to calculate the displacements and distortion of tunnels under different excavation sequences and soil reinforcement measures. Real-time monitoring confirms that the vertical displacements and diametrical distortion of tunnels are primarily caused by the excavation of the river above the tunnels, while horizontal displacements are induced by the excavation next to the tunnels. The study recommends a combination of cement slurry with a portal form and concrete with a plate form for soil reinforcement and tunnel protection.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Yaosheng Liu, Ang Li, Feng Dai, Ruochen Jiang, Yi Liu, Rui Chen
Summary: In this study, a hybrid model based on a multilayer perceptron (MLP) and meta-heuristic algorithms was developed to improve blast performance during tunnel excavation. Precise prediction of post-blasting indicators was important for optimization, and a comparison of meta-heuristic algorithms was conducted to find the most suitable model. The results showed that the developed model effectively reduces overbreak areas and quantitatively analyzes the influence of geological conditions.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Xiang Shen, Yifan Chen, Liqiang Cao, Xiangsheng Chen, Yanbin Fu, Chengyu Hong
Summary: In this paper, a machine learning-based method for predicting the slurry pressure in shield tunnel construction is proposed. By considering the influence of fault fracture zones and setting the formation influence coefficient, the accuracy of the prediction is significantly improved.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Shuying Wang, Zihao Zhou, Xiangcou Zheng, Jiazheng Zhong, Tengyue Zheng, Changhao Qi
Summary: A real-time assessment and monitoring approach based on laser scanning technology and point cloud data analysis was proposed to address the hysteresis in assessing the workability of conditioned soils and the inefficiency in estimating the soil volume flow rate in tunnelling practice. The approach was successfully applied in identifying the workability of conditioned soil and its discharge rate in the EPB shield tunnelling project.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Peng Jiang, Benchao Liu, Yuting Tang, Zhengyu Liu, Yonghao Pang
Summary: This study introduces a novel deep learning-based electrical method that jointly inverses resistivity and chargeability to estimate water-bearing structures and water volume. Compared with traditional linear inversion methods, the proposed method demonstrates superiority in locating and delineating anomalous bodies, reducing solution multiplicity.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Haoyu Mao, Nuwen Xu, Zhong Zhou, Chun Sha, Peiwei Xiao, Biao Li
Summary: The study focuses on the delineation of rock mass damage zones and stability analysis of underground powerhouse in Lianghekou hydropower station. ESG monitoring system is used to monitor the inner micro-fracture activity of surrounding rock mass in real-time. Engineering analogy method is adopted to forecast the deformation period of surrounding rock mass and analyze the variation characteristics of seismic source parameters. The research results provide references for similar deep underground excavation engineering in terms of design and construction.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Junling Qiu, Dedi Liu, Kai Zhao, Jinxing Lai, Xiuling Wang, Zhichao Wang, Tong Liu
Summary: This study focuses on the construction surface cracks of large cross-section tunnels in loess strata of China. The mechanism of surface crack formation is analyzed, and factors such as settlement deformation, construction scheme, and surrounding soil environment are identified as the main contributors. Numerical simulations were conducted to gain a deeper understanding of the influence of factors on surface cracks in loess tunnel construction. Specific measures for prevention and treatment of construction surface cracks are proposed to provide new ideas for surface crack control in loess tunnels.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Ting Shang, Jiaxin Lu, Ying Luo, Song Wang, Zhengyu He, Aobo Wang
Summary: The study reveals significant variations in car-following behavior across different types of tunnels and consecutive sections of the same tunnel. As tunnel length increases, the driving stability of following vehicles decreases, but the level of driving safety risk is not positively correlated with tunnel length. Significant vehicle trajectory oscillation is observed within the inner sections of long and extra-long tunnels, and a significant relationship between the acceleration of following vehicles and the location within the tunnel section is found.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Rusi Zeng, Zhongwei Shen, Jun Luo
Summary: The urban underground complexes (UUCs) in China have been effective in solving urban problems, but users have expressed dissatisfaction with the internal physical environment. Personal characteristics and environmental factors play significant roles in determining users' satisfaction.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Gabriel Lehmann, Heiko Kaeling, Sebastian Hoch, Kurosch Thuro
Summary: Analysing and predicting the advance rate of a tunnel boring machine (TBM) in hard rock is important for tunnelling projects. This study focuses on small-diameter TBMs and their unique characteristics, such as insufficient geotechnical information and special machine designs. A database of 37 projects with 70 geotechnically homogeneous areas is compiled to investigate the performance of small-diameter TBMs. The analysis shows that segment lining TBMs have higher penetration rates, and new approaches for the penetration prediction of pipe jacking machines in hard rock are proposed.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Ting Ren, Ming Qiao, Jon Roberts, Jennifer Hines, Yang-Wai Chow, Wei Zong, Adrian Sugden, Mark Shepherd, Matthew Farrelly, Gareth Kennedy, Faisal Hai, Willy Susilo
Summary: Long-term exposure to coal and silica dust during underground tunnelling operations is a growing concern. To bridge the gap between knowledge in dust exposure monitoring and frontline workers, a virtual reality educational tool was developed to visualize ventilation and dust flow characteristics. This tool allows workers to better understand decision-making and best practices for dust controls.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Dong Lin, Zhipeng Zhou, Miaocheng Weng, Wout Broere, Jianqiang Cui
Summary: Metro systems play a vital role in the transportation, economic, environmental and social aspects of cities. The uncertainties in construction, passenger comfort and safety, as well as efficiency and reliability of the metro system, have been widely studied. Metro systems influence urban development and have a positive impact on housing prices, public health, and environmental quality. Further research is needed to fill the research gaps and make recommendations for future studies.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)
Article
Construction & Building Technology
Wei Yu, Bo Wang, Xin Zi, Jie Dong
Summary: In this study, a whole-process analytical theory for the coupled deformation of deep circular tunnel surrounding rock and prestressed yielding anchor bolt (cable) system is derived and validated through numerical simulations. The results show that anchor bolts (cables) can significantly reduce the convergence of surrounding rock, and factors such as support timing and anchor cable length have important effects on the support effectiveness.
TUNNELLING AND UNDERGROUND SPACE TECHNOLOGY
(2024)