Study of historical evacuation drill data combining regression analysis and dimensionless numbers

Area of Science:

• Building safety engineering
• Human behavior in emergencies
• Statistical modeling

Background:

• Building evacuation time is influenced by numerous factors, including occupant behavior and architectural design.
• Existing research often lacks comprehensive analysis across diverse building types and large participant numbers.

Purpose of the Study:

• To analyze historical evacuation drill data to predict building evacuation efficiency.
• To develop a predictive model for the ratio between evacuation time and the number of people evacuated.
• To establish a framework for comparing the evacuation performance of different university buildings.

Main Methods:

• Analysis of historical data from 47 evacuation drills involving over 19,000 individuals across 15 university buildings.
• Application of dimensionless analysis to normalize evacuation parameters.
• Utilization of statistical regression models to establish relationships between evacuation time and occupant load.

Main Results:

• A predictive model was developed to estimate the ratio of exit time to the number of people evacuated.
• The dimensionless analysis and statistical regression approach proved effective in analyzing evacuation data.
• Significant correlations were found between building characteristics, occupant behavior, and evacuation times.

Conclusions:

• The proposed methodology offers a valuable tool for comparing the evacuation efficiency of different university buildings.
• This research highlights a promising avenue for future studies in building evacuation dynamics.
• The findings can be extended to assess and improve evacuation strategies in various building types.