Network redesign for efficient crowd flow and evacuation

Abstract

Over the past few years, many crowd related accidents have happened at places of mass gatherings such as religious places, venues of sports, political and social events, and railway stations due to poor crowd management or poor system design. To mitigate the impact of such tragedies, developing adequate crowd management strategies is essential. A critical component of crowd management strategies in such emergencies is an evacuation plan. A proactive measure for evacuation strategies is to make a plan to utilize the available capacity by efficient network redesign. Redesigning the network helps in influencing the demand by enhancing or limiting the use of certain routes, without the use of force or explicit guiding mechanisms. A bi-level network based optimization model is proposed for network redesign. This approach aims at an optimal capacity change for an efficient crowd flow and evacuation during mass gatherings. To determine the optimal capacity change required to achieve the desired evacuation state, an Interior Point Algorithm (IPA) is applied. Further, to increase the computational efficiency, a heuristic is developed that finds a good starting point for IPA. Finally, the robustness of the redesigned network with respect to evacuation demand is examined along with alternate measures that further help in reducing the evacuation time.