作者: Jan Faigl , Tomas Krajnik , Vojtech Vonasek , Libor Preucil
DOI: 10.1109/ICRA.2012.6224706
关键词:
摘要: This paper presents a multi-goal path planning framework based on self-organizing map algorithm and model of the navigation describing evolution localization error. The combines finding sequence goals' visits with goal-to-goal considering uncertainty. approach is able to deal local properties environment such as expected visible landmarks usable for navigation. affect performance navigation, therefore, can take full advantage information together global find improving autonomous Experimental results in real outdoor indoor environments indicate that provides paths effectively decreases uncertainty; thus, increases reliability visits.