This paper deals with the implementation of a coverage path planning algorithm for automated exploration of a known 2D world. The coverage path planning is the determination of set of goal points in the map, the robot must take in order to cover the entire environment. In this paper, segmentation based Boustrophedon Coverage algorithm is implemented and it is found that segmentation based implementation has certain advantages over classical implementation of Boustrophedon algorithm. The classical implementation is based on connectivity of a vertical sweeping line for determining obstacles in the environment. The Boustrophedon Decomposition is basically a generalization of the Trapezoidal Decomposition that allows non-polygonal obstacles as well. The Boustrophedon Decomposition generates far less cells as compared to Trapezoidal decomposition and hence improves the efficiency of coverage. The sensor based motion is embedded for safe navigation through unmapped obstacles. Application of coverage algorithms includes de-mining, vacuuming,sea-floor mapping, inspection, and space search. The platform used is Pioneer P3DX robot with a Hokuyo laser. The experimental setup is based on creation of a real world map and taking advantage of path planning algorithms provided by ARIA. Almost all the desired targets were successfully achieved and experiments are done on simulator and real robot.
展开▼