The sequences of bundles of line segments for autonomous robots with limited vision range to escape from blind alley regions

Abstract

Consider the following problem: A robot operating in a 2D environment with a limited vision range finds a path to a goal in an unknown environment containing obstacles. In this paper, we propose a novel algorithm to solve the problem. In some special cases, our algorithm is convergent with respect to $\Vert .\Vert$. The problem involves discovering the environmental map and blind alley regions, that are bounded by obstacles, and it provides no possible passage for robots except in and out of their path entry occur, the robot has to return back to some positions outside to escape from such regions such that the returned path is not longer than the path entry (Blind Alley Region problem, (BAR) problem, in short). To solve the (BAR) problem, sequences of bundles of line segments during the robot’s traveling are constructed in our algorithm.

Some advantages of our algorithm are that (a) It reduces search space in blind alley regions because it only works on the sequences of bundles of the line segments built by the robot’s limited vision range. (b) Our algorithm ensures that the returned path to escape from the regions is not longer than the previous path of the robot. (c) Due to the construction of the sequences of bundles of line segments, our paths are not always “close” obstacles and the number of turns of such paths is smaller ones determined by other shortest path algorithms (e.g., A*, RRT*).

Our algorithm is implemented in Python and we experience the algorithm on some autonomous robots with different vision ranges in real environment. We also compare our result with RRTX, a state-of-art local path-planning algorithm, and A${}^{\ast }$, a basic one. The experimental results show that our algorithm provides better solutions than RRTX and A* results in some specific circumstances.