Abstract
Creating complex and realistic crowd behaviors, such as pedestrian navigation behavior with dynamic obstacles, is a difficult and time consuming task. In this paper, we study one special type of crowd which is composed of urgent individuals, normal individuals, and normal groups. We use three steps to construct the crowd simulation in dynamic environment. The first one is that the urgent individuals move forward along a given path around dynamic obstacles and other crowd members. An optimal acceleration-velocity-bounded trajectory planning method is utilized to model their behaviors, which ensures that the durations of the generated trajectories are minimal and the urgent individuals are collision-free with dynamic obstacles (e.g., dynamic vehicles). In the second step, a pushing model is adopted to simulate the interactions between urgent members and normal ones, which ensures that the computational cost of the optimal trajectory planning is acceptable. The third step is obligated to imitate the interactions among normal members using collision avoidance behavior and flocking behavior. Various simulation results demonstrate that these three steps give realistic crowd phenomenon just like the real world.
Highlights
Crowds, ubiquitous in the real world from groups of humans to flocks of insects, are vital features to model in a virtual environment
We study one special type of crowd which is composed of urgent individuals, normal individuals, and normal groups
This paper focuses on one special crowd phenomenon which is significant in real world
Summary
Ubiquitous in the real world from groups of humans to flocks of insects, are vital features to model in a virtual environment. (1) Construct the behaviors of the urgent members with an optimal acceleration-velocity-bounded trajectory planning method This novel method ensures that the urgent members reach their destinations as soon as possible and are collision-free with the dynamic obstacles. A novel pushing model is adopted to ensure that the normal members will not get in the way of the urgent ones In this way, the normal members will not be treated as dynamic obstacles which promotes that the computational cost of the trajectory planning is acceptable for real time application. The crowd behaviors consist of four parts: optimal trajectory planning, pushing model, collision avoidance behavior, and flocking behavior, each of which is responsible for dealing with different interactions. After performing experiments with 3D virtual environments, simulation results and conclusions are stated at the end of this paper
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
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
