Comparison of Collision Avoidance Algorithms for Autonomous Multi-agent Systems

Autonomous multi-agent systems are raising in popularity in recent years. More specifically, Unmanned Aerial Vehicles (UAVs) are involved in modern solutions for surveillance, delivering and film shooting. To carry out these tasks, the avoidance of any possible collision is a crucial matter, mostly when agents need to cooperate. In this paper, different collision avoidance algorithms are compared and analyzed for distributed multi-agent holonomic systems. Our purpose is to identify and clarify the different classes of reciprocal collision avoidance algorithms and then to compare them using meaningful metrics and test for the evaluation.