Motion planning for constrained mobile robots in unknown environments

Abstract

The research presented in the thesis develops a framework for motion planning in unknown or partially known environments for mobile robots subject to various constraints. Both purely-sensor-based and map-aided approaches, combined with different strategies, are considered to accomplish global convergence to the goal, the primary aim of path planning tasks. Robot dynamic constraints and requirement of smooth motions are addressed in path planning (where the major concern is to find a smooth path satisfying various robot constraints) or motion planning (where the major concern is to generate an optimized, waypoint-directed feasible motion command satisfying certain objective functions). The research work include i) a practical approach of boundary following for a mobile robot with limited sensing ability; ii) a sensor-based hybrid path planning approach using polar polynomial curve for smooth, feasible path generation with collision test carried out in real-time; and iii) a hierarchical framework which robustly searches for an optimal global path incrementally based on a partial map updated periodically, and subsequently performs situation-dependant optimized dynamic motion planning for the robot to move at a relatively high speed while effectively avoiding collision with obstacles. The proposed solutions are demonstrated theoretically and empirically.