Crowd Evacuation Guidance Based on Combined Action Reinforcement Learning

Yiran Xue,Rui Wu,Jiafeng Liu,Xianglong Tang

doi:10.3390/a14010026

Abstract

Existing crowd evacuation guidance systems require the manual design of models and input parameters, incurring a significant workload and a potential for errors. This paper proposed an end-to-end intelligent evacuation guidance method based on deep reinforcement learning, and designed an interactive simulation environment based on the social force model. The agent could automatically learn a scene model and path planning strategy with only scene images as input, and directly output dynamic signage information. Aiming to solve the “dimension disaster” phenomenon of the deep Q network (DQN) algorithm in crowd evacuation, this paper proposed a combined action-space DQN (CA-DQN) algorithm that grouped Q network output layer nodes according to action dimensions, which significantly reduced the network complexity and improved system practicality in complex scenes. In this paper, the evacuation guidance system is defined as a reinforcement learning agent and implemented by the CA-DQN method, which provides a novel approach for the evacuation guidance problem. The experiments demonstrate that the proposed method is superior to the static guidance method, and on par with the manually designed model method.

Highlights

While large-scale shopping malls, office buildings, and other multi-functional buildings meet diverse needs, the complexity of buildings has gradually increased
As the original deep Q network (DQN) method would require 210 = 1024 output layer nodes for the experiments conducted compared with the 20 nodes for combined actionspace DQN (CA-DQN), the DQN
This paper proposed a crowd evacuation method based on combined action space deep reinforcement learning to overcome the disadvantages of existing methods

Summary

Introduction

While large-scale shopping malls, office buildings, and other multi-functional buildings meet diverse needs, the complexity of buildings has gradually increased. When disasters such as earthquakes and fires occur, the complex structures of buildings hinder evacuation and create a new safety threat. It is difficult for crowds to identify an optimal escape route, owing to people’s ignorance of the building environment, their limited vision, and them panicking. Under the influence of herd behavior, survivors are prone to cause congestion, or even trampling, risking significant additional loss of life [1]. A method for guiding crowd evacuation using the most effective route is of great significance for protecting lives and reducing personal and property losses during disasters. Researchers have developed several crowd evacuation guidance systems based on dynamic guidance signs [1,2,3,4] to assist crowds to evacuate effectively during a disaster

Methods

Results

Conclusion