Liangyu Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Liangyu Li has a total of 35 co-authors. Here are the top co-authors by number of shared publications:
Chenchen Wang
2 shared publicationsHai Shao
School Belarusian State University 220004 Minsk Belarus
2 shared publicationsJiajun Jin
2 shared publicationsShiyi Xu
School Belarusian State University 220004 Minsk Belarus
2 shared publications- Xiaoyan Zhao
Faculty of Information Science & Technology, Universiti Kebangsaan Malaysia, Bangi Selangor, Malaysia
2 shared publications - Xuwei Guo
Universiti Kebangsaan Malaysia, Malaysia
2 shared publications
Latest Publications
- 30 October 2026
- Published by Springer Nature
- p. 62-74
AI Adoption and FinTech Advantages of Large Firms: Triangulated Evidence from Europe
- 30 October 2026
- Published by Springer Nature
- p. 62-74
AI Adoption and FinTech Advantages of Large Firms: Triangulated Evidence from Europe
- Open Access
- 24 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 733-737
Research on Digital Twin and Multi-Agent Collaborative Systems in Smart Manufacturing
The integration of Digital Twin (DT) technology and Multi-Agent Systems (MAS) has attracted increasing attention as an effective approach for enabling distributed intelligence and collaborative decision-making in smart manufacturing. Digital twins provide real-time virtual representations of physical systems, while multi-agent systems support decentralized control and coordination in complex industrial environments. However, existing studies on DT–MAS integration are scattered and lack a systematic understanding of research trends and system architectures. This paper presents a bibliometric and architecture-oriented review of digital twin–enabled multi-agent collaborative systems in smart manufacturing. Based on 324 publications retrieved from the Web of Science Core Collection from 2017 to 2026, publication trends, collaboration networks, and keyword co-occurrence patterns are analyzed to identify major research themes. Furthermore, representative studies are examined to summarize typical DT–MAS architecture patterns, including centralized, distributed, and hybrid configurations. Key technical challenges and future development directions are discussed to support the design of scalable and intelligent manufacturing systems.
- 24 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 733-737
Open AccessResearch on Digital Twin and Multi-Agent Collaborative Systems in Smart Manufacturing
The integration of Digital Twin (DT) technology and Multi-Agent Systems (MAS) has attracted increasing attention as an effective approach for enabling distributed intelligence and collaborative decision-making in smart manufacturing. Digital twins provide real-time virtual representations of physical systems, while multi-agent systems support decentralized control and coordination in complex industrial environments. However, existing studies on DT–MAS integration are scattered and lack a systematic understanding of research trends and system architectures. This paper presents a bibliometric and architecture-oriented review of digital twin–enabled multi-agent collaborative systems in smart manufacturing. Based on 324 publications retrieved from the Web of Science Core Collection from 2017 to 2026, publication trends, collaboration networks, and keyword co-occurrence patterns are analyzed to identify major research themes. Furthermore, representative studies are examined to summarize typical DT–MAS architecture patterns, including centralized, distributed, and hybrid configurations. Key technical challenges and future development directions are discussed to support the design of scalable and intelligent manufacturing systems.
- Open Access
- 1 January 2026
- Published by Institution of Engineering and Technology (IET) in IET Software
- Vol. 2026 (1)
Configuration Factors for Compact LLM‐Based Story Point Estimation: A Controlled Project‐Level Empirical Study
Story point estimation is widely used in agile software planning, yet automated estimation remains challenging because issue descriptions and story point distributions vary substantially across projects. This study examines configuration sensitivity and comparative performance in compact large language model (LLM)‐based story point estimation through a controlled project‐level study using five TAWOS projects and two ~1–1.5B‐parameter compact backbones with frozen parameters and trainable prediction heads. The study comprises four‐factor configuration analysis covering maximum sequence length, metadata concatenation, learning rate, and soft‐target mixing; metadata‐free sensitivity analysis; learning‐rate–soft‐target‐mixing interaction analysis; comparison with three aligned baselines; assessment of alternative temperature‐selection objectives; and computational‐cost evaluation. Soft‐target mixing and learning rate show the larger average mean absolute error (MAE) effects among the four evaluated factors, with mean high‐minus‐low effects of −0.0535 and −0.0458, respectively, although both vary in direction across project–backbone–seed scenarios. The metadata‐free analysis shows similar average effect magnitudes for learning rate and soft‐target mixing, while the interaction between these two factors is heterogeneous, with 11 negative and 9 positive scenario‐level contrasts. When compact configurations and baseline methods are selected independently using seed‐averaged validation MAE, the corresponding test results favor the compact configuration in three projects and the selected baseline in two, with additional project‐specific variation among TFIDF‐LGBM, MiniLM‐MLP, and CodeBERT‐FT. Temperature‐selection rules produce different MAE, snapped‐expected calibration error (ECE), and negative log‐likelihood (NLL) profiles. The compact configurations require very small trainable prediction heads but retain substantially higher inference latency and memory requirements than the evaluated baselines; method‐level mean raw‐text‐to‐logits latency ranges ~0.37–25.72 ms/sample in the reported NVIDIA A10 environment. Overall, the results characterize compact‐LLM story point estimation by heterogeneous configuration sensitivity, project‐dependent model ranking, and accuracy–cost trade‐offs rather than a single consistently preferable configuration or model family.
- 1 January 2026
- Published by Institution of Engineering and Technology (IET) in IET Software
- Vol. 2026 (1)
Open AccessConfiguration Factors for Compact LLM‐Based Story Point Estimation: A Controlled Project‐Level Empirical Study
Story point estimation is widely used in agile software planning, yet automated estimation remains challenging because issue descriptions and story point distributions vary substantially across projects. This study examines configuration sensitivity and comparative performance in compact large language model (LLM)‐based story point estimation through a controlled project‐level study using five TAWOS projects and two ~1–1.5B‐parameter compact backbones with frozen parameters and trainable prediction heads. The study comprises four‐factor configuration analysis covering maximum sequence length, metadata concatenation, learning rate, and soft‐target mixing; metadata‐free sensitivity analysis; learning‐rate–soft‐target‐mixing interaction analysis; comparison with three aligned baselines; assessment of alternative temperature‐selection objectives; and computational‐cost evaluation. Soft‐target mixing and learning rate show the larger average mean absolute error (MAE) effects among the four evaluated factors, with mean high‐minus‐low effects of −0.0535 and −0.0458, respectively, although both vary in direction across project–backbone–seed scenarios. The metadata‐free analysis shows similar average effect magnitudes for learning rate and soft‐target mixing, while the interaction between these two factors is heterogeneous, with 11 negative and 9 positive scenario‐level contrasts. When compact configurations and baseline methods are selected independently using seed‐averaged validation MAE, the corresponding test results favor the compact configuration in three projects and the selected baseline in two, with additional project‐specific variation among TFIDF‐LGBM, MiniLM‐MLP, and CodeBERT‐FT. Temperature‐selection rules produce different MAE, snapped‐expected calibration error (ECE), and negative log‐likelihood (NLL) profiles. The compact configurations require very small trainable prediction heads but retain substantially higher inference latency and memory requirements than the evaluated baselines; method‐level mean raw‐text‐to‐logits latency ranges ~0.37–25.72 ms/sample in the reported NVIDIA A10 environment. Overall, the results characterize compact‐LLM story point estimation by heterogeneous configuration sensitivity, project‐dependent model ranking, and accuracy–cost trade‐offs rather than a single consistently preferable configuration or model family.
- Open Access
- 31 October 2024
- Published by Atlantis Press SARL
- p. 455-460
The Impact of Social Media on Television Broadcasting and Film Production
- 31 October 2024
- Published by Atlantis Press SARL
- p. 455-460
Open AccessThe Impact of Social Media on Television Broadcasting and Film Production
- Open Access
- 20 October 2024
- Published by Atlantis Press SARL
- p. 83-89
The Integration of Computer Engineering and Economic Sustainability Forecasting in Smart Manufacturing
- 20 October 2024
- Published by Atlantis Press SARL
- p. 83-89
Open AccessThe Integration of Computer Engineering and Economic Sustainability Forecasting in Smart Manufacturing
- Top Cited Papers
- 17 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Consumer Electronics
- Vol. 70 (3) , 5401-5410
Harnessing Multimodal Data and Deep Learning for Comprehensive Gait Analysis in Pediatric Cerebral Palsy
Cerebral palsy (CP) is a leading cause of motor dysfunction in children, significantly impacting gait and mobility. Accurate and early diagnosis of gait abnormalities in pediatric CP patients is crucial for effective intervention and management. However, making an early-stage CP diagnosis based only on a single vision modality such as an MRI has many difficulties. Because of the baby’s obstinate movements, the possibility of early recovery, the lack of a single vision modality, and the noisy or absent brain magnetic resonance imaging (MRI) slices, the task is getting harder and harder. This study employed a robust framework that leverages data from multiple sensor modalities, including wearable inertial measurement units (IMUs), pressure-sensitive mats, and motion capture systems integrated with MRI to generate multimodal data. This multimodal data was then processed using convolutional neural networks (CNNs) and long short-term memory (LSTM) networks to capture both spatial and temporal dynamics of gait patterns. In the experimentation, we achieved remarkable results with an accuracy of 95.33%, an AUC of 96.2%, an F1 score of 95.28%, and a misclassification rate of 0.0467. Also, the comparative analysis with state-of-the-art demonstrates that the proposed approach significantly outperforms traditional methods in identifying subtle gait abnormalities, providing a more detailed and accurate assessment of gait deviations in pediatric cerebral palsy patients.
- 17 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Consumer Electronics
- Vol. 70 (3) , 5401-5410
Top Cited PapersHarnessing Multimodal Data and Deep Learning for Comprehensive Gait Analysis in Pediatric Cerebral Palsy
Cerebral palsy (CP) is a leading cause of motor dysfunction in children, significantly impacting gait and mobility. Accurate and early diagnosis of gait abnormalities in pediatric CP patients is crucial for effective intervention and management. However, making an early-stage CP diagnosis based only on a single vision modality such as an MRI has many difficulties. Because of the baby’s obstinate movements, the possibility of early recovery, the lack of a single vision modality, and the noisy or absent brain magnetic resonance imaging (MRI) slices, the task is getting harder and harder. This study employed a robust framework that leverages data from multiple sensor modalities, including wearable inertial measurement units (IMUs), pressure-sensitive mats, and motion capture systems integrated with MRI to generate multimodal data. This multimodal data was then processed using convolutional neural networks (CNNs) and long short-term memory (LSTM) networks to capture both spatial and temporal dynamics of gait patterns. In the experimentation, we achieved remarkable results with an accuracy of 95.33%, an AUC of 96.2%, an F1 score of 95.28%, and a misclassification rate of 0.0467. Also, the comparative analysis with state-of-the-art demonstrates that the proposed approach significantly outperforms traditional methods in identifying subtle gait abnormalities, providing a more detailed and accurate assessment of gait deviations in pediatric cerebral palsy patients.
- Open Access
- 13 March 2024
- Published by Darcy & Roy Press Co. Ltd. in Highlights in Science, Engineering and Technology
- Vol. 85, 1314-1322
Biomedical Engineering Research: Computer Translating and Teaching System and Neural Network
In the research on medical students, the research team obtained the idea of using English teaching to cultivate interdisciplinary talents, and established a computer teaching model to predict the incidence rate of cancer through the use of computer engineering based in-depth learning and oncology. Through the situation of susceptible gene mutation and computer imaging after chromosome staining, an algorithm teaching model using convolutional neural network and linear function is proposed. This model solves the problem that computer engineering students lack knowledge of medicine. This model can enhance medical students' understanding of applied mathematics and neural networks, provide new ideas for the medical community to comprehensively promote in-depth learning, and design social experiments, carry out data analysis and data visualization. The research team analyzed and reported this situation, summarized and communicated.
- 13 March 2024
- Published by Darcy & Roy Press Co. Ltd. in Highlights in Science, Engineering and Technology
- Vol. 85, 1314-1322
Open AccessBiomedical Engineering Research: Computer Translating and Teaching System and Neural Network
In the research on medical students, the research team obtained the idea of using English teaching to cultivate interdisciplinary talents, and established a computer teaching model to predict the incidence rate of cancer through the use of computer engineering based in-depth learning and oncology. Through the situation of susceptible gene mutation and computer imaging after chromosome staining, an algorithm teaching model using convolutional neural network and linear function is proposed. This model solves the problem that computer engineering students lack knowledge of medicine. This model can enhance medical students' understanding of applied mathematics and neural networks, provide new ideas for the medical community to comprehensively promote in-depth learning, and design social experiments, carry out data analysis and data visualization. The research team analyzed and reported this situation, summarized and communicated.
- Open Access
- 1 January 2023
- Published by European Alliance for Innovation n.o.
Research and Economic Analysis of Education Information System Based on Intelligent Economic Product Research and Development -Model Design and Economic Benefit Analysis of Teaching Information System in Colleges and Universities Based on Artificial Intelligence Course, Hunan Medical University As An Example
We have designed the computer system of teaching science popularization for physical network, virtual reality technology, augmented reality technology, artificial neural network and other models, with the purpose of improving the learning quality and learning situation of students targeted by teaching. At the same time, we have tried to put forward the strategy of global e-commerce network for various computer teaching models, with the purpose of solving the communication problem of language conversion through computer intelligence technology, At the same time, the research team discussed various kinds of colleges and universities. During the discussion, the research team put forward the modeling strategy of computer teaching system for students of different majors, and we conducted economic benefit analysis. The research team analyzed the model design, data analysis, data processing and discussion results, and conducted global communication, which is reported in this article.
- 1 January 2023
- Published by European Alliance for Innovation n.o.
Open AccessResearch and Economic Analysis of Education Information System Based on Intelligent Economic Product Research and Development -Model Design and Economic Benefit Analysis of Teaching Information System in Colleges and Universities Based on Artificial Intelligence Course, Hunan Medical University As An Example
We have designed the computer system of teaching science popularization for physical network, virtual reality technology, augmented reality technology, artificial neural network and other models, with the purpose of improving the learning quality and learning situation of students targeted by teaching. At the same time, we have tried to put forward the strategy of global e-commerce network for various computer teaching models, with the purpose of solving the communication problem of language conversion through computer intelligence technology, At the same time, the research team discussed various kinds of colleges and universities. During the discussion, the research team put forward the modeling strategy of computer teaching system for students of different majors, and we conducted economic benefit analysis. The research team analyzed the model design, data analysis, data processing and discussion results, and conducted global communication, which is reported in this article.
- Open Access
- 1 January 2023
- Published by European Alliance for Innovation n.o.
Pedagogy Research Based on Artificial Intelligence Technology and Digital Economy Product R&D-Research on education information system based on virtual reality technology, taking Hunan Medical University of China as an example
We have studied the research and development of digital economy products by using artificial intelligence technology. Its core is to use virtual reality technology to solve the problem of functional design of educational information system. At the same time, we have put forward an innovative computer model and combined the application program of computer client to integrate the entertainment function into the overall educational information system. This strategy can solve the problem of students' enthusiasm for learning. At the same time, we analyzed Hunan Medical University in China and conducted research. Through the research results, we completed the market economics prospect analysis of the education information system. The core of the digital economy product is market oriented. Therefore, the research team conducted a sociological experiment. The research team currently analyzes and reports the engineering model design and digital economy product development, as well as the experimental results and data, and discusses them at the end.
- 1 January 2023
- Published by European Alliance for Innovation n.o.
Open AccessPedagogy Research Based on Artificial Intelligence Technology and Digital Economy Product R&D-Research on education information system based on virtual reality technology, taking Hunan Medical University of China as an example
We have studied the research and development of digital economy products by using artificial intelligence technology. Its core is to use virtual reality technology to solve the problem of functional design of educational information system. At the same time, we have put forward an innovative computer model and combined the application program of computer client to integrate the entertainment function into the overall educational information system. This strategy can solve the problem of students' enthusiasm for learning. At the same time, we analyzed Hunan Medical University in China and conducted research. Through the research results, we completed the market economics prospect analysis of the education information system. The core of the digital economy product is market oriented. Therefore, the research team conducted a sociological experiment. The research team currently analyzes and reports the engineering model design and digital economy product development, as well as the experimental results and data, and discusses them at the end.