Automatical Guardrail Design of Subway Stations through Multi-objective Evolutionary Algorithm

In subway stations, elevators are one of the most narrowed areas that slow down the moving of crowds. A large number of passengers gather around the elevator entrances and may cause unexpected accidents such as stampede. An effective way to guide the flow of passengers is to use guardrails. So far, the arrangement of guardrails in most subway stations is still designed manually, which requires rich experience and expert knowledge. In this paper, we propose to use the multi-objective evolutionary algorithm to design the guardrails of the elevator entrance automatically. The transfer time of passengers and the flow rate are optimized concurrently. The proposed algorithm is tested in two scenarios with different complexities. Experimental results show that the proposed algorithm can provide promising guardrail arrangements, and reveal some instructive conclusions for guardrail design in subway stations.