| 引用本文: | 张学习,杨宜民.基于多传感器信息融合的移动机器人快速精确自定位[J].控制理论与应用,2011,28(3):443~448.[点击复制] |
| ZHANG Xue-xi,YANG Yi-min.Fast and accurate self-localization of mobile robot based on multi-sensor[J].Control Theory and Technology,2011,28(3):443~448.[点击复制] |
|
| 基于多传感器信息融合的移动机器人快速精确自定位 |
| Fast and accurate self-localization of mobile robot based on multi-sensor |
| 摘要点击 2580 全文点击 3528 投稿时间：2009-07-13 修订日期：2010-06-03 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 10.7641/j.issn.1000-8152.2011.3.CCTA090919 |
| 2011,28(3):443-448 |
| 中文关键词 移动机器人 自定位 数据融合 粒子滤波 |
| 英文关键词 mobile robot location data fusion particle filter |
| 基金项目 国家自然科学基金–广东省自然科学基金联合基金资助项目(U0735003); 广东省“211工程”资助项目(粤发改[431]). |
|
| 中文摘要 |
| 通过分析全向视觉、电子罗盘和里程计等传感器的感知模型, 设计并实现了一种给定环境模型下移动机器人全局自定位算法. 该算法利用蒙特卡罗粒子滤波, 融合多个传感器在不同观测点获取的观测数据完成机器人自定位. 与传统的、采用单一传感器自定位的方法相比, 它把多个同质或异质传感器所提供的不完整测量及相关联数据库中的信息加以综合, 降低单个信息的不确定性, 提高了自定位的精度. 同时由于充分利用了全向视觉传感器的观测模型, 在粒子滤波定位方法的实现过程中, 粒子点的置信度通过查表的方法获得, 达到了在自定位过程中, 确定置信度时所要求的快速性, 从而保证了定位算法较好的实时性. 实验结果证明了该方法的有效性. |
| 英文摘要 |
| A method based on the fusion of multi-sensor information is proposed for self-localization of the mobile robot in known environment. It provides the configuration of the multi-sensor information fusion system and analyzes the all-forward wheel, omni-vision and electrical compass. The Monte Carlo(MCL) particle filtration method combines the measured data of sensors in various observation points to achieve the fusion localization for the mobile robot. Being different from the traditional single-sensor self-localization of the mobile robot, this method synthesizes the incomplete information from heterogeneous or homogeneity sensors and the related data from the data-base, thus, reducing the uncertainty in the information from a single sensor and improving the accuracy in self-localization. Because of the use of the observation model of the omni-vision, and the employment of the lookup table for determining the confidence interval in the realization of self-localization, this method achieves the required rapidity in obtaining the confidence interval for self-localization, ensuring its application in real time. Experimental results validate the proposed method. |
|
|
|
|
|