研究火灾发生时楼梯等地点容易产生拥堵的地方的人流研究，给出最佳路径-FireEscape A Genetic Programming Approach to the Optimization of Building Evacuations



Contribute to the FireEscape program now



Abstract Many evacuation routes are drawn without considering traffic flow patterns. As a result, bottlenecks can occur when large numbers of people converge in areas such as narrow hallways, doors, and intersections. FireEscape is an appliation designed to consider traffic flow and generate an optimal route. This paper describes the specific methods and techniques used to create the program. Using modern computing techniques such as: genetic programming, distributed computing, and cellular automation we attempt to simulate evacuations and develop an efficient set of routes to vacate the premises. Modeling the complex and chaotic movements of individuals in an evacuation is accomplished by simulating the logic used by real people when navigating through densely populated zones. Optimization of the multivariable set through genetic programming yields a sol