Collaborative Optimal Reciprocal Collision Avoidance (CORCA)

Abstract

Mobile robots are the future of modern day vehicles and their collision is one of the major concerns and challenges which is a very active eld of research these days. As the number of vehicles increasing, the complexity and challenges are also growing exponentially. There is an imminent need of a strategy or al- gorithm to mitigate their problems. In that regard, the research presented in this thesis is one of possible solutions that eliminates the drawback of existing collision avoidance algorithms known as Deadlock. The proposed algorithm is named as Velocity-based Collaborative Optimal Reciprocal Collision Avoidance (V-CORCA) which is extensively analyzed using holonomic robots in a computer simulation. A brief comparison among other versions of the same techniques such as Optimal Reciprocal Collision Avoidance (ORCA) and Collaborative Optimal Reciprocal Collision Avoidance (CORCA), is also presented for performance eval- uation. The simulation results reveal that unlike ORCA, V-CORCA was able to prevent deadlocks, while at the same time, reducing the overall halt time for single and multiple robots within an eight robots-based multi-robot system. By analyzing the simulation results of a single and (eight robots-based) multi-robot system , it is concluded that V-CORCA is not only e cient in reducing the over all halt time but also eradicates the possibility of occurring a Deadlock.