We propose a motion planning gap-based algorithms for mobile robots in an unknown environment\nfor exploration purposes. The results are locally optimal and sufficient to navigate and explore\nthe environment. In contrast with the traditional roadmap-based algorithms, our proposed\nalgorithm is designed to use minimal sensory data instead of costly ones. Therefore, we adopt a\ndynamic data structure called Gap Navigation Trees (GNT), which keeps track of the depth discontinuities\n(gaps) of the local environment. It is incrementally constructed as the robot which navigates\nthe environment. Upon exploring the whole environment, the resulting final data structure\nexemplifies the roadmap required for further processing. To avoid infinite cycles, we propose to\nuse landmarks. Similar to traditional roadmap techniques, the resulting algorithm can serve key\napplications such as exploration and target finding. The simulation results endorse this conclusion.\nHowever, our solution is cost effective, when compared to traditional roadmap systems, which\nmakes it more attractive to use in some applications such as search and rescue in hazardous environments.
Loading....