Application of a coordinated trajectory planning and real-time obstacle avoidance algorithm

Abstract

A novel method for coordinated trajectory planning and real-time obstacle avoidance of autonomous systems is presented. The desired autonomous system trajectories are generated from a set of first order ODEs. The solution to this system of ODEs converges to either a desired target position or a closed orbit defined by a limit cycle. Coordinated control is achieved by utilizing the nature of limit cycles where independent, non-crossing paths are automatically generated from different initial positions that smoothly converge to the desired closed orbits. Real-time obstacle avoidance is achieved by specifying a transitional elliptically shaped closed orbit around the nearest obstacle blocking the path. This orbit determines an alternate trajectory that avoids the obstacle. When the obstacle no longer blocks a direct path to the original target trajectory, a transitional trajectory that returns to the original path is defined. The coordination and obstacle avoidance methods are demonstrated experimentally using differential-drive wheeled mobile robots.