基于图像和人工鱼群算法的建筑火灾动态疏散路径规划研究

为了解决大型建筑发生火灾时传统静态疏散系统无法根据火灾点和人员拥挤程度进行路径调整这一问题，提出了基于图像和人工鱼群算法的动态疏散路径规划方法。在栅格图上进行路径规划，通过将鱼群的最优解替换为可行解，使鱼群避免陷入局部最优和全局最优相互干扰的情况，并结合摄像头采集图像，通过人脸识别人数，判断当前路径是否拥挤，及时调整路径，从而确保规划出的路径可以避免堵塞，动态疏散人群。仿真实验结果表明，所提算法能够在相同时间内，规划出较蚁群算法路径更短，可避免陷入局部最优和死锁状态，根据拥挤程度及时改变路径，并能够在时间和空间双重约束的情况下实现人群动态疏散。因此，新算法在相同运行时间内可以规划出更短的路径，可以帮助火灾现场人群以更少时间、更短路径、更高效率的方式进行动态疏散。

Research on dynamic evacuation path planning of building fire based on image and artificial fish swarm algorithm

In order to solve the problem that the traditional static evacuation system can not adjust the path according to the fire point and personnel congestion in the case of large-scale building fire, a dynamic evacuation path based on image and artificial fish swarm algorithm (AFSA) was proposed. The algorithm carried out path planning on the grid map. By replacing the optimal solution of the fish swarm with the feasible solution, the fish swarm could avoid the interference between the local optimal and global optimal. Combined with the images collected by the camera, it could judge whether the current path was crowded or not through face recognition, and adjust the path in time, so as to ensure that the planned path can avoid congestion and evacuate the crowd dynamically. Experimental results show that the algorithm can plan a shorter path than ant colony algorithm in the same time, avoid falling into local optimization and deadlock state, change the path in time according to the degree of congestion, and realize dynamic evacuation under the double constraints of time and space. In this study, shorter path can be planned within the same running time, so it the crowd dynamic evacuation with less time, shorter path and higher efficiency.