A local evaluation approach for multi-agent navigation in dynamic scenarios.

In recent years, the technology for crowd simulation has been applied in many fields. However, collision avoidance considering of multiple individuals and moving obstacles simultaneously is still a challenging task in this research area. In this paper, we present a novel technique for multi-agent navigation in dynamic scenario. By coupling unified representation of environment with a agent-based evaluation model, our method takes into account dynamic and static environment conditions simultaneously. Each individual make an estimation of the costs-to-moving and perform a balanced decision to react to multiple requests. Moreover, our agent-based evaluation approach provides similar operation for each agent. Therefore, we can make full use of the processing capacity of GPU with this parallel characteristic. The experimental results show that the algorithm can depict the interactions between virtual agents and dynamic environments. Also thousands of agents can be simulated in real-time.