Multi-Objective Optimization Framework for Emergency Building Sweep Operations: A Mixed-Integer Programming Approach with Uncertainty

Abstract

Emergency evacuation efficiency directly impacts survival rates during fire incidents. Statistics indicate that survival probability decreases by 7–10% per minute after eight min- utes without rescue. We develop a multi-objective route optimization framework balancing sweep completion time and occupant waiting duration. The baseline Mixed Integer Lin- ear Programming (MILP) model minimizes weighted makespan and waiting costs across homogeneous settings. For complex scenarios involving heterogeneous occupants, we in- tegrate a Fuzzy Membership Degree model handling type uncertainty. The framework extends to multi-floor buildings with responder volume optimization, validated through progressive scenarios from simple office layouts to complex college buildings. An Ant Colony Optimization algorithm combined with local search efficiently solves large-scale instances. Extensions incorporate dynamic injury deterioration in fire emergencies and drone-assisted inspection. Sensitivity analysis confirms model robustness, with makespan remaining stable across weight variations. Results demonstrate that layout complexity significantly impacts required personnel, with responder needs increasing from 2 to 13 as building complexity and dynamic efects intensify.