Robust Range-Only SLAM for Unmanned Aerial Systems

详细信息    查看全文

• 作者：Felipe R. Fabresse ; Fernando Caballero…
• 关键词：Simultaneous localization and mapping ; Range ; only ; Unmanned aerial systems ; Extended Kalman filter
• 刊名：Journal of Intelligent & Robotic Systems
• 出版年：2016
• 出版时间：December 2016
• 年：2016
• 卷：84
• 期：1-4
• 页码：297-310
• 全文大小：
• 刊物类别：Engineering
• 刊物主题：Control, Robotics, Mechatronics; Electrical Engineering; Artificial Intelligence (incl. Robotics); Mechanical Engineering;
• 出版者：Springer Netherlands
• ISSN：1573-0409
• 卷排序：84

文摘

Range-only sensors, in contrast to bearing-only sensors or LIDAR sensors are sensors which only provides distance between two points but not bearing information. The reduced amount of information provided by this sensors makes their application in Simultaneous Localization and Mapping (SLAM) problems specially challenging due to the lack of information. The complexity of the problem becomes higher when using noisy range-only sensors like radio-based where the multi-path effect may introduce range outliers, hence reducing the precision of the SLAM solution. Furthermore, the application of Range-only SLAM to aerial systems implies a higher level of complexity due to the introduction to a 3th dimensionality in contrast to ground mobile robots where the altitude dimension is not necessary in most applications. This paper presents an efficient and robust method to the Range-only SLAM (RO-SLAM) problem when applied to aerial vehicles and radio-based range-only measurements. The method is based on a pre-filtering algorithm which is able to detect and remove range outliers and on the application of a reduced parameterization which scales better with 3D RO-SLAM problems. The method is validated with real experiments involving an aerial vehicle and multiple radio-based range-only sensors.