Tianjin University你好, Hello, こんにちは, Bonjour, Привет, Hallo, مرحبا, नमस्ते, 안녕하세요, Ciao. I'm Liu Zenghui. I am currently completing my master's studies at the School of Future Technology, Tianjin University, primarily engaged in fire dynamics, building fire analysis, and intelligent fire protection. The supervisor of my master's program is Professor Gang Liu. Between October 2022 and July 2023, I conducted one-year research on the spatiotemporal prediction of infection risks in waiting rooms as a research assistant. In the future, I will devote myself to research on resilience, sustainability, and equity of complex urban systems. Starting from September 2026, I will pursue a doctoral degree in Civil Engineering at The University of British Columbia. My research focus will be on national-scale building decarbonization, urban energy systems, and life cycle assessment. I am also a student member of the Architectural Society of China, the Fire Engineers Association, the International Society for Urban Information, and the International Building Performance Simulation Association. I also serve as a peer reviewer for journals such as Humanities and Social Sciences Communications, Science Progress, Knowledge-Based System, and Journal of Safety Science and Resilience. Since April 2025, I have been on probation as a young editorial board member of Jandoo press Journal of Sustainable Built Environment (JSBE). In my spare time, I play basketball, table tennis, Guan Dan (a kind of card game), listen to music, and play the guitar (for indie music lovers).
") does not match the recommended repository name for your site ("").
", so that your site can be accessed directly at "http://".
However, if the current repository name is intended, you can ignore this message by removing "{% include widgets/debug_repo_name.html %}" in index.html.
",
which does not match the baseurl ("") configured in _config.yml.
baseurl in _config.yml to "".

Zenghui Liu,Mingyu Zhai,Weiyao Yang,Jing Lin,Yao Sun*.
Developments in the Built Environment 2025
We used machine learning to analyze over 100,000 residential fire incidents in the UK and identified the true drivers of fire spread. The results showed that the most crucial factor was the emergency response stage - such as delayed alarm, time of discovery, etc., accounting for approximately half of the risk impact. The second was the fire dynamics characteristics (such as rapid growth, kitchen fires), accounting for about 30%. The attributes of buildings and people overall have a significant weight, but they can amplify the risk in specific situations. Based on this, resources should be prioritized towards "earlier detection + faster response": optimizing the deployment of fire-fighting forces, upgrading alarm/detection technologies, and building intelligent monitoring platforms to improve the efficiency of early warning and response.
Zenghui Liu,Mingyu Zhai,Weiyao Yang,Jing Lin,Yao Sun*.
Developments in the Built Environment 2025
We used machine learning to analyze over 100,000 residential fire incidents in the UK and identified the true drivers of fire spread. The results showed that the most crucial factor was the emergency response stage - such as delayed alarm, time of discovery, etc., accounting for approximately half of the risk impact. The second was the fire dynamics characteristics (such as rapid growth, kitchen fires), accounting for about 30%. The attributes of buildings and people overall have a significant weight, but they can amplify the risk in specific situations. Based on this, resources should be prioritized towards "earlier detection + faster response": optimizing the deployment of fire-fighting forces, upgrading alarm/detection technologies, and building intelligent monitoring platforms to improve the efficiency of early warning and response.

GANG Liu#,Zenghui Liu#,Guanhua Qu*,Lei Ren*.
Process Safety and Environmental Protection 2024
This study combines distributed fiber optic temperature sensing systems with deep learning algorithms to develop a dual-agent intelligent fire detection method for rapidly and accurately predicting key fire information in large commercial spaces, including the location of the fire source, the intensity of fire development, and the distribution of carbon monoxide on critical planes. This work not only improves the accuracy and response speed of fire alarms but also provides data support for emergency evacuation and rescue operations, reducing casualties and property loss caused by fires, and has significant practical importance for the safety protection systems of modern commercial buildings.
GANG Liu#,Zenghui Liu#,Guanhua Qu*,Lei Ren*.
Process Safety and Environmental Protection 2024
This study combines distributed fiber optic temperature sensing systems with deep learning algorithms to develop a dual-agent intelligent fire detection method for rapidly and accurately predicting key fire information in large commercial spaces, including the location of the fire source, the intensity of fire development, and the distribution of carbon monoxide on critical planes. This work not only improves the accuracy and response speed of fire alarms but also provides data support for emergency evacuation and rescue operations, reducing casualties and property loss caused by fires, and has significant practical importance for the safety protection systems of modern commercial buildings.

Guanhua Qu#,Zenghui Liu#,Lei Ren*,Gang Liu*.
Journal of Building Engineering 2024
The innovation of this study is to generate infection risk map by intelligent method and manage the system, so as to achieve accurate control of airborne infection risk in waiting room. The results of this study not only help to select hospital design scheme, but also optimize the operation control strategy of ventilation system, so as to strengthen the control of infectious diseases and reduce the threat of cross-infection of respiratory infectious diseases to public health safety during peak periods in hospitals..
Guanhua Qu#,Zenghui Liu#,Lei Ren*,Gang Liu*.
Journal of Building Engineering 2024
The innovation of this study is to generate infection risk map by intelligent method and manage the system, so as to achieve accurate control of airborne infection risk in waiting room. The results of this study not only help to select hospital design scheme, but also optimize the operation control strategy of ventilation system, so as to strengthen the control of infectious diseases and reduce the threat of cross-infection of respiratory infectious diseases to public health safety during peak periods in hospitals..

Zenghui Liu*,Yingnan Zhuang.
Journal of Building Engineering 2024
This study not only improves the emergency response strategies for urban residential fires, but also provides customized fire safety policies for different urban environments, effectively reducing fire risk. Through the introduction of resampler technology and interpretable machine learning, this study not only improves the prediction accuracy of the model, but also enhances the transparency and interpretability of the model, providing a more scientific and reliable tool for fire risk management..
Zenghui Liu*,Yingnan Zhuang.
Journal of Building Engineering 2024
This study not only improves the emergency response strategies for urban residential fires, but also provides customized fire safety policies for different urban environments, effectively reducing fire risk. Through the introduction of resampler technology and interpretable machine learning, this study not only improves the prediction accuracy of the model, but also enhances the transparency and interpretability of the model, providing a more scientific and reliable tool for fire risk management..