Abstract

Controlled experiments involved more than 700 participants inside a metro train model were conducted to investigate effects of voice alarms on the evacuation process. The process was divided into 4 stages from the perspective of collective behavior: the recognition stage (from the emergency happens to the first person standing up), the response stage (from the first person standing up to the last person standing up), the transfer stage (from the last person standing up to the first person leaving the carriage), and the queue stage (from the first person leaving the carriage to the last person leaving the carriage). Results showed that the mean evacuation time of all experiments was 77.61 s, and mean percentages of the four stages were 16.89%, 12.80%, 30.67% and 30.76%, respectively. When focusing on a specific stage, a shorter voice alarm cycle, a closer voice source and a more rapid tone resulted in a shorter recognition time. In the response stage, the response order was mainly affected by physical distance. A closer and more directive voice guided evacuees to find the exit more quickly, and further reduced the transfer time. Opening doors by using emergency devices took about 75% of the transfer time on average. Voice alarm didn’t play a key role in the queue stage, and the queue time was mainly affected by the number and the width of available doors. The results provide references for developing models and designing Public Announcement system of a metro train.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call