Faster Calculations and Intelligent Ventilation Control for Tunnel Fires through Artificial Intelligence


  • Yao Hong
  • Ceji Fu (supervisor)
  • Bart Merci (supervisor)


This research project aims to develop: (1) a rapid calculator for predicting the tunnel fire dynamics, (2) an intelligent ventilation method for suppressing the spread of tunnel fires. As a significant infrastructure hazard, fires in tunnels cause a great deal of economic and social losses each year. Predicting the behavior of fires rapidly is very important for the timely rescuing of lives and properties. However, a considerable amount of computing time is required because of the enormous computing domain (e.g. in a 10 km tunnel). Methodologies are needed to accelerate tunnel fire prediction will be used as possible. Artificial intelligence will be adopted to divide the different regions of the tunnel, so as to ensure calculation accuracy and speed. High-performance computing technology will be used to accelerate the calculations near the fire. In this way, a complete calculation system for faster tunnel fire simulation will be established, which can not only ensure the accuracy of capturing the flow field near the fire source but also accelerate the calculation. Finally, the system will be used to study the intelligent ventilation method, where the AI and cybernetics methods will be used to design a fully automatic ventilation control scheme. In general, this study may help to save time for tunnel fire prediction, so as to increase the speed and possibility of rescue. Moreover, an automatic ventilation scheme will also help to reduce the consequences of fire.



Funding CSC Junyi Li