Optimization of elevator evacuation considering potential over-crowding

Many countries, such as the US, China, and the UK, have considered using safe elevators (such as fire elevators) on safe floors (such as refuge floors) in high-rise buildings to speed up emergency evacuation. However, for safety reasons such as sheave traction and brake capability, elevators are designed not to operate when overloaded. Overcrowding, then, may actually slow down building evacuation when a car is overloaded. As a result, how to speed up elevator evacuation considering potential overcrowding is an important issue. This problem is difficult because of the uncertain traffic and complicated elevator scheduling. To reduce the traffic uncertainties, one trend is to use the surveillance cameras in smart high-rise buildings to know the number of waiting evacuees. To avoid the phenomenon of elevator overloading, multiple elevators will be sent to the same floor based on the number of waiting evacuees. To reduce the complication of elevator scheduling, all the elevators are divided into groups with one or more elevators in each group. Then these groups will be dispatched to different floors according to the numbers of waiting evacuees. Our objective is to minimize the evacuation time with the constraints of evacuees´ safety. The Genetic Algorithm toolbox in Matlab is used to solve this problem. A case study is used to demonstrate that, compared with the building where one floor can be served by one car, our elevator scheduling can reduce the evacuation time.