中文摘要：

针对人群疏散的仿真模拟中,如何引导虚拟人群避免与障碍物发生碰撞,顺利通过复杂环境到达期望目标的问题,提出一种基于语义信息的路线规划方法。该方法首先采用语义信息实现全局路径规划,进而利用势场方法来避免碰撞,最终完成人群的快速疏散。仿真结果表明,该方法在产生高质量路径的同时,可以快速灵活地完成人群在复杂场景中的运动,提高了疏散效率。

英文摘要：

In the simulation of crowd evacuation,how to guide the virtual crowd through complex environment and to avoid collision with obstacles to reach the desired goal is a problem. We put forward a kind of route planning method based on semantic information.This method firstly realizes global path planning using semantic information,and then uses the potential field method to avoid collision with obstacles,and finally completes the rapid evacuation of crowd. Simulation results show that the method can produce high-quality path and guide the virtual crowd movement quickly and flexibly in complex scenes to improve the efficiency of evacuation.