A Comparative Study of Velocity Obstacle Approaches for Multi-Agent Systems

Abstract

This paper presents a critical analysis of some of the most promising approaches aimed at geometrically generating reactive avoidance trajectories for multi-agent systems. Several evaluation scenarios are proposed that include both sensor uncertainty and increasing difficulty. An intensive 1000 cycle Monte Carlo analysis is used to assess the performance of the selected algorithms under the presented conditions. The Optimal Reciprocal Collision Avoidance (ORCA) method was shown to demonstrate the most scalable computation times and collision likelihood in the presented scenarios. The respective features and limitations of the algorithms are discussed and presented through examples.