Abstract
Distance potential field is useful method for pedestrian/crowd/evacuation simulations or robotics. In simulation, distance potential field enables agents to find their paths toward destinations effectively. The simulations are more important in large-scale spaces than small-scale spaces. Because area size and interior structure are large and complex, in large-scale spaces, occupant's moving distance and path are distant and complicated. Also, phenomenons that caused by a lot of occupants are considerably complex. In large-scale space, however, calculation time of distance potential field increases markedly. To improve problems of existing methods, we developed Hierarchical Distance Potential Field. That has dual scale models and calculate distance potential field combining hierarchically them. Showing sample spaces, we demonstrated that's concept and algorithm. Improvements of Hierarchical Distance Potential Field are as follows : 1) Dividing large-scale space to tiny parts, calculation time is little than existing methods. 2) When distance potential field must be calculated numerously, this method can processes them at one time. 3) Using divided tiny parts, this method provides multi-threading calculation. If computing environments allow, it can maximize them.
