Crowd evacuation navigation for evasive maneuver of brownian based dynamic obstacles using reciprocal velocity obstacles

Susi Juniastuti, Moch Fachri, Fresy Nugroho, Supeno Mardi Susiki Nugroho, Mochamad Hariadi

Abstract


This paper presents an approach for evasive maneuver against dynamic obstacles in multi-agent navigation in a crowd evacuation scenario. Our proposed approach is based on reciprocal velocity obstacles (RVO) with a different manner to treat the obstacles. We treat all possible hindrances in velocity space reciprocally thus all collision cones generated by other agents and obstacles are treated in the same RVO manner with the key difference in the effort of avoidance. Our approach assumes that dynamic obstacles bear no awareness of navigation space unlike agents thus the avoidance effort lies on behalf of the mobile agents, creating unmutual effort in an evasive maneuver. We display our approach in an evacuation scenario where a crowd of agents must navigate through an evacuation area trespassing zone filled with dynamic obstacles. These dynamic obstacles consist of random motion built based on Brownian motion thus posses an immense challenge for the mobile agent in order to overcome this hindrance and safely navigate to their evacuation area. Our experimentation shows that 51.1% fewer collisions occurred which is denote safer navigation for agents in approaching their evacuation point.


Keywords


Brownian motion; Collision avoidance; Crowd evacuation; Mobile agents; Reciprocal velocity obstacles

Full Text:

PDF


DOI: https://doi.org/10.11591/eei.v11i4.3806

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Bulletin of EEI Stats