| URN | etd-0825125-191418 | Statistics | This thesis had been viewed 15 times. Download 0 times. |
| Author | Wei-Ting Chung | ||
| Author's Email Address | 523weiting@gmail.com | ||
| Department | Institute of construction technology | ||
| Year | 2024 | Semester | 2 |
| Degree | Master | Type of Document | Master's Thesis |
| Language | zh-TW.Big5 Chinese | Page Count | 57 |
| Title | Research on Constructing Regional Fire Risk Using AIˇGTaking the New Taipei City Metropolitan Area as an Example | ||
| Keyword | |||
| Abstract | New Taipei City has a complex urban structure, consisting of older neighborhoods, high-density residential areas, and various industrial plants. Furthermore, it faces extreme weather events brought on by climate change, such as drought and high temperatures, which can increase the risk of fire. To address these challenges and reduce fire risk, this study, drawing on the New Taipei City Fire Department's safety management system and incorporating technology from startups, employed innovative technologies and methods to develop a prototype "Fire Risk-Based Prediction Model." This model predicts the spatial distribution of high and potentially catastrophic fire risks, thereby enhancing fire prevention and response capabilities. Traditional fire risk management methods mainly rely on historical data analysis, manual experience judgment and regular fire safety inspections. It is difficult to achieve objective and accurate risk assessment, requires a lot of manpower and material resources, and cannot focus on monitoring high-risk areas. The "Fire Risk-Oriented Prediction Model" integrates multiple sources of data, including fire safety inspection records, geographic information, demographics, and meteorology. Through big data analysis and machine learning, the New Taipei City building risk model and risk map were created. This model accurately predicts fire risks in different areas, helping the Fire Department prioritize limited resources in high-risk areas and achieve targeted fire prevention and inspections. Simulation tests conducted in this study found that the New Taipei City building risk model achieved an accuracy rate of 77%, accurately predicting the level of fire risk in various locations. In the future, property management units can use this risk prediction to prioritize fire inspections and fire prevention awareness programs in high-risk areas, providing a reference for decision-making. |
||
| Advisor Committee | |||
| Files | indicate access worldwide | ||
| Date of Defense | 2025-07-15 | Date of Submission | 2025-08-25 |