《计算机应用研究》|Application Research of Computers

传感器数据融合与单一传感器在同步定位与构图中的对比研究

Comparative study of sensor data fusion and single sensor in simultaneous localization and mapping

免费全文下载 (已被下载 次)  
获取PDF全文
作者 郭维,彭辉,张瑜
机构 成都信息工程大学 a.软件工程学院;b.控制工程学院,成都 610225
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2020)06-058-1882-04
DOI 10.19734/j.issn.1001-3695.2018.11.0914
摘要 针对室内环境下单一传感器数据在同步定位与构图中的局限性问题,提出了一种使用模糊贴近度和自适应加权相结合的传感器数据融合方法。首先建立了观测模型,并使用模糊贴近度的方法将Kinect深度摄像头获取的数据和激光雷达数据进行首次融合;然后使用自适应加权的方法对权值系数进行二次加权修正;最后将使用该方法构建的地图与单一传感器构建的地图进行了对比分析。实验证明,使用该方法获得的数据更加可靠,获取的地图精确度更高、环境特征表现更加平滑,验证了所提算法的有效性。
关键词 传感器数据融合; 模糊贴近度; 自适应加权; 同步定位与构图
基金项目 国家自然科学基金资助项目(61472050)
四川省科技厅资助项目(2019YJ0356)
本文URL http://www.arocmag.com/article/01-2020-06-058.html
英文标题 Comparative study of sensor data fusion and single sensor in simultaneous localization and mapping
作者英文名 Guo Wei, Peng Hui, Zhang Yu
机构英文名 a.School of Software Engineering,b.School of Control Engineering,Chengdu University of Information Technology,Chengdu 610225,China
英文摘要 Aiming at the limitation of single sensor data in simultaneous localization and mapping in indoor environment, this paper proposed a sensor data fusion method using fuzzy nearness and adaptive weighting. Firstly, it established the observation model, and firstly merged the data acquired by the Kinect depth camera and the LiDAR data by the method of fuzzy nearness. Then it used the weighting coefficient for the second weighting correction by the method of adaptive weighting. Finally, it compared the map built by this method with the map built by a single sensor. The experimental results show that the data obtained by this method is more reliable, and the obtained maps have higher accuracy and smoother environmental features, which verifies the effectiveness of the proposed algorithm.
英文关键词 sensor data fusion; fuzzy nearness; adaptive weighting; SLAM
参考文献 查看稿件参考文献
 
收稿日期 2018/11/27
修回日期 2019/1/22
页码 1882-1885
中图分类号 TP391.9;TP399
文献标志码 A