| 基于粒子滤波联合算法的地磁室内定位 |
| |
| 引用本文： | 黄鹤,仇凯悦,李维,罗德安. 基于粒子滤波联合算法的地磁室内定位[J]. 西南交通大学学报, 2019, 54(3): 604-610. DOI: 10.3969/j.issn.0258-2724.20170927 |
| |
| 作者姓名： | 黄鹤 仇凯悦 李维 罗德安 |
| |
| 作者单位： | 北京建筑大学测绘与城市空间信息学院;北京建筑大学北京未来城市设计高精尖创新中心 |
| |
| 基金项目： | 国家重点研发计划资助项目（2017YFB0503702） |
| |
| 摘 要： | 室内地磁场受钢结构与其它铁磁材料的影响，造成磁场区域局部异常，使室内地磁场具有特异性. 受益于此种现象，室内地磁定位技术得以实现. 然而在大型建筑中地磁场的特异性会减弱，这导致定位结果出现模糊现象. 针对这一现象，文中提出了基于路径匹配的室内地磁定位技术，通过增加匹配特征数量来解决此问题. 使用基于动态时间规整（dynamic time warp，DTW）算法与粒子滤波（particle filter，PF）算法的新型联合算法，并以路径匹配的模式对目标进行追踪. 在匹配过程中又通过计算斯皮尔曼等级（Spearman）相关系数确定路径之间的相似度，使之作为辅助定位. 最后用装载了磁传感器的测量机器人进行实验验证，结果表明:路径匹配具有足够的地磁特征数量，能够解决特异性减弱情况下定位结果模糊现象，且定位精度优于1 m.
|
| 关 键 词： | 粒子滤波 地磁场 动态时间规整 室内定位系统 目标跟踪 |
| 收稿时间： | 2018-01-09 |
Indoor Geomagnetic Positioning Based on Joint Algorithm of Particle Filter |
| |
| HUANG He,QIU Kaiyue,LI Wei,LUO Dean. Indoor Geomagnetic Positioning Based on Joint Algorithm of Particle Filter[J]. Journal of Southwest Jiaotong University, 2019, 54(3): 604-610. DOI: 10.3969/j.issn.0258-2724.20170927 |
| |
| Authors: | HUANG He QIU Kaiyue LI Wei LUO Dean |
| |
| Abstract: | The indoor geomagnetic field is affected by the presence of a steel structure and other ferromagnetic materials, causing local anomalies in the magnetic field region and and a unique indoor magnetic field. Indoor geomagnetic positioning technology can be realized by exploiting this phenomenon. However, in large buildings, the specificity of the geomagnetic field is weakened, which leads to the distortion of positioning results. To overcome this problem, an indoor geomagnetic positioning technology based on path matching was proposed, which increases the number of matching features. A new joint algorithm was used combining the dynamic time warp algorithm with the particle filter algorithm, which can track the target in a path matching mode. In the matching process, the similarity between the paths was determined by calculating the Spearman correlation coefficient, to assist the positioning. Finally, experimental verification was performed using a measuring robot loaded with a magnetic sensor. The results show that the path matching exhibits a sufficient number of geomagnetic features, which can overcome the phenomenon of blurring the positioning results under the condition of weakening specificity, and the positioning accuracy is higher than 1 m. |
| |
| Keywords: | |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《西南交通大学学报》浏览原始摘要信息 |
|
点击此处可从《西南交通大学学报》下载免费的PDF全文 |
|
