Intelligent navigation in a dynamic virtual environment

Intelligent virtual human is widely applied in educational, entertainment, and simulation software. In this paper, a three-level control model is presented to realize smart navigation in a dynamic environment. It improves the subsumption architecture by adding the ability of global planning. Collision avoidance and path planning are realized in the 0-level layer and the 1-level layer, respectively. The simple expert system in the 2-level layer realizes more intelligent navigation. Finally, a experiment was done in our human animation platform. And the result shows that virtual human with this control model can navigate collision-freely in the dynamic environment.