Date of this Version
Decentralized Collision Avoidance, M.S Thesis, Jayasri K Janardanan, 2013, University of Nebraska
Autonomous Robots must carry out their tasks as independently as possible and each robot may be assigned different tasks at different locations. As these tasks are being performed, the robots have to navigate correctly such that the assigned tasks are completed efficiently, while also avoiding each other and other obstacles. To accomplish effective navigation, we must ensure that the robots are calibrated to avoid colliding with any kind of object on its path. Each robot has to sense the obstacles on its path and take necessary corrective measure to avoid those obstacles. In a situation with multiple robots, robots may cross each other’s paths and thus algorithms have to be developed to ensure collision avoidance among them.
Collision avoidance among multiple robots has been studied extensively over many years. In this thesis, we investigate the Reciprocal n-body Collision Avoidance Algorithm (RCAA) where collision avoidance among multiple robots is addressed. One advantage of RCAA over other techniques is the decentralized approach that allows robots to take collision- avoidance decisions by themselves using only velocity and position of the nearby robots that are along its trajectory. Though this method is widely used, a major limitation is the assumption of perfect sensing, which is not a guaranteed behavior in real environments. In real world scenarios, erroneous measurements may be obtained during which the RCAA is not capable of ensuring perfect collision avoidance. This limitation in the RCAA needs to be addressed and in this thesis, we have devised a method to address this using particle filters.
A particle filter is appended to the RCAA to sample velocities and thereby provide the robots with more options to avoid coming in the path of each other. A simulation program has been developed to implement the entire system showcasing different scenarios where the introduction of particle filter has made the system more stable as it ensures a more streamlined and efficient collision avoidance among robots.
Adviser: Carrick Detweiler