Article
Computer Science, Artificial Intelligence
Yucheng Wang, Fei Tao, Ying Zuo, Meng Zhang, Qinglin Qi
Summary: In this paper, a fault diagnosis method based on digital twin enhanced signed directed graph (SDG) is proposed for autoclave. The SDG model of the temperature control system is constructed and improved by using digital twin. Experimental results show that the proposed method can improve the speed and resolution of fault diagnosis.
JOURNAL OF INTELLIGENT MANUFACTURING
(2023)
Article
Chemistry, Multidisciplinary
Yuzhan Dong, Qing Chen, Wei Ding, Ning Shao, Guanting Chen, Guanbin Li
Summary: This paper studies the application of digital twin technology in real-time state analysis of protection systems. By constructing a state evaluation model and a fault prediction model, the effectiveness and accuracy of the method are verified based on actual operation data, providing technical support for the operation and maintenance of protection systems.
APPLIED SCIENCES-BASEL
(2022)
Article
Engineering, Multidisciplinary
Jian Zheng, Dezhi Jiang, Xuan Jia, Cong Wang, Qingfeng Zhang, Frantisek Brumercik, Zhixiong Li
Summary: In the era of significant data, green power is seen as the future of ship development. Establishing fast-response marine power models to evaluate different green power-driven components is crucial. However, evaluating a finite model often requires costly re-meshing when model parameters change. The use of reduced-order models can address this issue by rapidly calculating model responses, though real-time computing remains challenging. This study introduces a digital twin reduced-order model for marine centrifugal pumps, demonstrating its accuracy and feasibility for fault detection techniques.
ENGINEERING ANALYSIS WITH BOUNDARY ELEMENTS
(2023)
Article
Ecology
Athanasios Trantas, Ruduan Plug, Paolo Pileggi, Elena Lazovik
Summary: Digital Twin is a contemporary digital representation paradigm that encompasses complex interactions in the natural environment. Developing Digital Twin Applications for biodiversity can uncover anthropogenic effects causing biodiversity loss and identify ways to mitigate or prevent these effects. However, there are unique challenges in applying Digital Twin to biodiversity, such as heterogeneous modeling, model co-simulation, variable computational power, and integration with existing research infrastructures.
ECOLOGICAL INFORMATICS
(2023)
Article
Engineering, Industrial
Litong Zhang, Yu Guo, Weiwei Qian, Weili Wang, Daoyuan Liu, Sai Liu
Summary: This paper proposes a modelling and online training method for digital twin workshop to address the difficulties in modelling, simulation, and verification. It describes a multi-level digital twin aggregate modelling method and a digital twin organization system. A spatio-temporal data model is constructed based on the data demand for digital twin aggregates. The paper also presents a training method using truncated normal distribution and a verification method based on real-virtual error for digital twin models. The effectiveness of real-time status monitoring, online model training, and production simulation is verified through a case study.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2023)
Article
Engineering, Multidisciplinary
Yucheng Wang, Fei Tao, Ying Zuo, Meng Zhang, Qinglin Qi
Summary: Composite materials are widely used due to their excellent properties, but quality defects can lead to lower quality components and potential accidents. Predicting the quality of composite materials accurately is difficult due to uncertain curing conditions and incomplete feature space. To address this, a digital twin visual model is constructed and coupled with a variable composite material model to simulate the curing process. Simulated data is used to enhance quality prediction, and an extreme learning machine is trained for prediction. The effectiveness of the method is verified through result analysis.
Article
Engineering, Mechanical
Chao Yang, Baoping Cai, Rui Zhang, Zhexian Zou, Xiangdi Kong, Xiaoyan Shao, Yiliu Liu, Haidong Shao, Javed Akbar Khan
Summary: A subsea production system is crucial for the production of oil and gas underwater. Real-time monitoring of the subsea production control system is essential for ensuring safety. This article proposes a methodology that combines digital twin technology and fault diagnosis to improve the accuracy of diagnosing minor faults in the subsea production control system.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Computer Science, Interdisciplinary Applications
Concetta Semeraro, Mario Lezoche, Herve Panetto, Michele Dassisti
Summary: The Digital Twin (DT) is a virtual copy of a physical system that predicts failures and opportunities for change, prescribes actions in real-time, and optimizes unexpected events. However, modeling the virtual copy is complex and requires accurate models. This paper proposes a new approach that uses modeling patterns and their invariance property to design a DT. The potential of invariance modeling patterns is demonstrated through a real industrial application.
JOURNAL OF INDUSTRIAL INFORMATION INTEGRATION
(2023)
Article
Construction & Building Technology
Haidar Hosamo Hosamo, Henrik Kofoed Nielsen, Dimitrios Kraniotis, Paul Ragnar Svennevig, Kjeld Svidt
Summary: This paper evaluates the performance of buildings in terms of occupant comfort using a probabilistic model based on Bayesian networks. It proposes a novel Digital Twin approach that combines building information modeling, real-time sensor data, occupants' feedback, and HVAC faults detection and prediction. The study also presents methods for using BIM as a visualization platform and for predictive maintenance of HVAC systems.
ENERGY AND BUILDINGS
(2023)
Article
Computer Science, Information Systems
Ravitej Bhagavathi, D. Kwame Minde Kufoalor, Agus Hasan
Summary: This paper presents a digital twin-driven fault diagnosis approach for autonomous surface vehicles. An adaptive extended Kalman filter algorithm is proposed to estimate the magnitude of the faults by calculating the parameter estimation gains directly from the sensor systems. The algorithm is tested in an autonomous surface vehicle called the Otter and is able to accurately detect and estimate actuator faults.
Review
Multidisciplinary Sciences
Dong Zhong, Zhelei Xia, Yian Zhu, Junhua Duan
Summary: The development of manufacturing industry requires predictive maintenance to be more effective, and digital twin technology has emerged as a promising solution. This paper introduces the general methods of digital twin technology and predictive maintenance, analyzes the gap between them, and emphasizes the importance of using digital twin technology for predictive maintenance. Furthermore, it presents the predictive maintenance method based on digital twin, discusses its characteristics and differences from traditional predictive maintenance, and showcases its applications in various industries.
Article
Engineering, Mechanical
Rituraj Rituraj, Rudolf Scheidl
Summary: Digital twins are becoming essential in mechanical systems for their advantages in performance analysis, predictive maintenance, and realtime monitoring. This article proposes a simple yet accurate model for counterbalance valves (CBVs) commonly used in hydraulic systems, synthesizing parameters for internal geometry, inertia, and friction characteristics. The model is validated through experiments and proves successful in predicting CBV behavior, making it suitable for integration in larger systems.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Engineering, Multidisciplinary
Haohan Tao, Peng Jia, Xiangyu Wang, Xi Chen, Liquan Wang
Summary: A digital twin-based framework is proposed for data-driven fault diagnosis in a subsea control system. A novel modeling technique, the physics informed temporal convolution network, is developed by combining physics-based simulation with collected sensor signals. The framework utilizes the digital twin to generate simulated signals for training a convolutional neural network-based diagnostic model. An online model modification technique is proposed to continuously train the model using real-time data. Experimental results demonstrate that the proposed framework outperforms traditional methods, particularly with limited labeled samples, and the online model modification technique improves diagnostic accuracy.
Article
Computer Science, Artificial Intelligence
Evgeny Zotov, Ashutosh Tiwari, Visakan Kadirkamanathan
Summary: Manufacturing digitalisation is a crucial aspect of transitioning to Industry 4.0, with digital twin playing a significant role. This paper introduces a digital twin component based on generative adversarial networks for production process optimization and simulation uncertainty estimation.
INTEGRATED COMPUTER-AIDED ENGINEERING
(2021)
Article
Computer Science, Hardware & Architecture
Bilgehan Erman, Catello Di Martino
Summary: This article discusses the vision of the 6G era and the concept of digital twin. Challenges remain in designing, delivering, and maintaining private wireless networks for autonomous industrial settings, including immediate detection of performance problems and use-case-dependent SLA compliance prediction. The article presents a solution that utilizes deep neural network models and generative adversarial network methods for continuous testing and SLA management of networks.
Article
Engineering, Industrial
Jingwei Guo, Ying Cheng, Dongxu Wang, Fei Tao, Stefan Pickl
Summary: Smart manufacturing is a popular concept that focuses on smarter decision-making and increasing production efficiency. This article proposes the use of Dataspace as a feasible and effective method to manage data in smart manufacturing. The contribution of this research is the design of an industrial Dataspace platform that accommodates the characteristics of smart manufacturing, such as harnessing distributed data, understanding industrial data through ontology or knowledge, and enabling related decisions. An analytical case in Surface Mounting Technology manufacturing is provided to demonstrate the customization and decision support offered by industrial Dataspace.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2023)
Article
Engineering, Industrial
Ying Cheng, Yanshan Gao, Lei Wang, Fei Tao, Qing-Guo Wang
Summary: This paper studies the operational robustness of the IIoT platform for Manufacturing Service Collaboration (MSC) and proposes a graph-based operational robustness analysis method. By modelling the IIoT platform operation network for MSC as an interdependent network structure, the effectiveness of MSC under uncertainty effects can be quantified.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2023)
Editorial Material
Computer Science, Interdisciplinary Applications
Yongkui Liu Lihui, Lihui Wang, Sotiris Makris, Jorg Krueger
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2023)
Article
Engineering, Multidisciplinary
Shibao Pang, Shunsheng Guo, Xi Vincent Wang, Lei Wang, Lihui Wang
Summary: An Industrial Internet platform is essential for smart manufacturing, allowing physical manufacturing resources to be virtualized and enabling collaboration between resources in the form of services. This paper proposes a dual-dimensional service collaboration methodology that combines functional and amount collaboration to optimize manufacturing service collaboration. A multi-objective memetic algorithm with multiple local search operators is tailored to solve the problem of service selection and amount distribution. Experimental results show that the algorithm outperforms commonly used algorithms in terms of convergence, solution quality, and comprehensive metrics.
Review
Engineering, Manufacturing
Chengxi Li, Pai Zheng, Yue Yin, Baicun Wang, Lihui Wang
Summary: To meet the demands of personalized smart manufacturing, Deep Reinforcement Learning (DRL) has gained significant attention as an adaptive and flexible solution. By combining the strengths of Deep Neural Networks (DNN) and Reinforcement Learning (RL), DRL enables precise and fast decision-making in dynamic and complex situations. However, there is currently a lack of comprehensive review on the application of DRL in smart manufacturing. This study addresses this gap by conducting a systematic review to provide insights into the development, application, and challenges of DRL in smart manufacturing, with a focus on the engineering lifecycle stages.
CIRP JOURNAL OF MANUFACTURING SCIENCE AND TECHNOLOGY
(2023)
Article
Social Issues
Qiuchen Wang, Hongyi Liu, Fredrik Ore, Lihui Wang, Jannicke Baalsrud Hauge, Sebastiaan Meijer
Summary: This study explores the implementation of industrial collaborative robots in the automotive manufacturing industry and the concerns of multiple actors involved. The research finds that the actors have positive expectations for the adoption of collaborative robots, but there are different concerns among actors from different organizations.
TECHNOLOGY IN SOCIETY
(2023)
Article
Engineering, Manufacturing
Yaoyao Ping, Yongkui Liu, Lin Zhang, Lihui Wang, Xun Xu
Summary: Cloud manufacturing is a service-oriented networked manufacturing model that aims to provide manufacturing resources as services in an on-demand manner. Scheduling is one of the key techniques for cloud manufacturing to achieve the aim. Deep reinforcement learning (DRL), specifically the Deep Q-network (DQN) algorithm, is found to be an effective approach for multitask scheduling in cloud manufacturing, outperforming traditional algorithms such as genetic algorithm (GA) and ant colony optimization (ACO) algorithm.
JOURNAL OF MANUFACTURING SCIENCE AND ENGINEERING-TRANSACTIONS OF THE ASME
(2023)
Article
Engineering, Industrial
Yiyuan Qin, Xianli Liu, Caixu Yue, Mingwei Zhao, Xudong Wei, Lihui Wang
Summary: In this paper, a tool wear recognition and prediction method based on stack sparse self-coding network is proposed for effective monitoring of tool wear in metal cutting process to ensure the machining quality of parts. This method simplifies the establishment process of monitoring model, monitors tool wear based on different task requirements, and guides tool replacement in the actual cutting process.
JOURNAL OF MANUFACTURING SYSTEMS
(2023)
Article
Engineering, Industrial
Wei Ma, Xianli Liu, Caixu Yue, Lihui Wang, Steven Y. Liang
Summary: This paper proposes a multi-scale one-dimensional convolution network model based on multi-model fusion learning for tool wear monitoring. The model extracts multi-scale local features of multi-sensor signals and uses multi-model fusion learning skills to improve prediction performance and adaptability to actual industrial environment.
JOURNAL OF MANUFACTURING SYSTEMS
(2023)
Editorial Material
Engineering, Industrial
Baicun Wang, Tao Peng, Xi Vincent Wang, Thorsten Wuest, David Romero, Lihui Wang
JOURNAL OF MANUFACTURING SYSTEMS
(2023)
Review
Engineering, Industrial
Andrea de Giorgio, Fabio Marco Monetti, Antonio Maffei, Mario Romero, Lihui Wang
Summary: This article presents a systematic literature review of XR technologies in manufacturing education, highlighting the importance of bridging the gap between theory and practice. It also provides a comprehensive overview of development tools and experimental strategies, serving as a guide for evaluating XR interventions in manufacturing education and training.
JOURNAL OF MANUFACTURING SYSTEMS
(2023)
Article
Engineering, Industrial
Kendrik Yan Hong Lim, Theresia Stefanny Yosal, Chun-Hsien Chen, Pai Zheng, Lihui Wang, Xun Xu
Summary: The increasing complexity of industrial systems requires more effective and intelligent maintenance approaches to address manufacturing defects. This paper introduces a cognitive digital twin system that leverages industrial knowledge graphs to support maintenance planning and operations. The system can manage interconnected systems, facilitate cross-domain analysis, and generate feasible solutions validated through simulation. It can also identify potential disruptions in new product designs.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2023)
Article
Computer Science, Artificial Intelligence
Bowen Zhang, Xianli Liu, Caixu Yue, Shaoyang Liu, Xuebing Li, Steven Y. Liang, Lihui Wang
Summary: This study proposes a novel tool condition monitoring (TCM) method based on data augmentation. It uses generative adversarial networks (GANs) to balance data distribution and combines continuous wavelet transform (CWT) with convolutional neural network (CNN) for monitoring. The results suggest that data augmentation can effectively improve monitoring performance.
JOURNAL OF INTELLIGENT MANUFACTURING
(2023)
Article
Computer Science, Interdisciplinary Applications
Yaqian Zhang, Kai Ding, Jizhuang Hui, Sichao Liu, Wanjin Guo, Lihui Wang
Summary: Human-robot collaborative assembly combines human flexibility and robot efficiency in mass personalization production. To improve robot's cognitive ability, this research proposes a two-stage skeleton-RGB integrated model for recognizing highly similar human actions, an online prediction approach for predicting human actions ahead of schedule, and a dynamic response scheme for accurate part positioning and continuous update of human actions. Experimental results demonstrate the effectiveness of the proposed model and approach in achieving precise human action recognition and online prediction with lower computational cost.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2024)
Review
Engineering, Industrial
Meng Zhang, Fei Tao, Ying Zuo, Feng Xiang, Lihui Wang, A. Y. C. Nee
Summary: Intelligent algorithms play a crucial role in smart manufacturing by providing optimal solutions to improve manufacturing processes. This paper comprehensively surveys and analyzes relevant literature to identify the top ten commonly used algorithms and studies their application issues and challenges. These findings are of great significance for the development and application of smart manufacturing.
JOURNAL OF MANUFACTURING SYSTEMS
(2023)
Article
Engineering, Industrial
Xiaoliang Yan, Reed Williams, Elena Arvanitis, Shreyes Melkote
Summary: This paper extends prior work by developing a semantic segmentation approach for machinable volume decomposition using pre-trained generative process capability models, providing manufacturability feedback and labels of candidate machining operations for query 3D parts.
JOURNAL OF MANUFACTURING SYSTEMS
(2024)
Article
Engineering, Industrial
Jing Huang, Zhifen Zhang, Rui Qin, Yanlong Yu, Guangrui Wen, Wei Cheng, Xuefeng Chen
Summary: In this study, a deep learning framework that combines interpretability and feature fusion is proposed for real-time monitoring of pipeline leaks. The proposed method extracts abstract feature details of leak acoustic emission signals through multi-level dynamic receptive fields and optimizes the learning process of the network using a feature fusion module. Experimental results show that the proposed method can effectively extract distinguishing features of leak acoustic emission signals, achieving higher recognition accuracy compared to typical deep learning methods. Additionally, feature map visualization demonstrates the physical interpretability of the proposed method in abstract feature extraction.
JOURNAL OF MANUFACTURING SYSTEMS
(2024)