Simulation of Passenger Ship Emergency Evacuation Based on Neural Network Algorithm and Physics Mechanical Model

Abstract

In recent years, the research on emergency evacuation of passenger ships has attracted much attention. The research on emergency evacuation mostly focuses on public places on land, while the research on emergency evacuation of passenger ships at sea is less. Based on the neural network algorithm, the social force model is improved by using the mechanical properties in physics, and a typical passenger ship personnel evacuation model is established by using the anylogic simulation software. There are significant differences in the passenger assembly time and congestion formation area under different evacuation modes. Therefore, the passenger ship personnel evacuation process should be analyzed in combination with the accident type and severity, This study provides theoretical basis and decision support for the design and evaluation of evacuation safety of large passenger ships.