基于惯性传感器和视觉里程计的机器人定位  

Robot localization algorithm based on inertial sensor and video odometry

作　　者：夏凌楠[1] 张波[2] 王营冠[1] 魏建明[2] 

机构地区：[1]中国科学院上海微系统与信息技术研究所无线传感网与通信重点实验室,上海200050 [2]中国科学院上海高等研究院,上海201210

出　　处：《仪器仪表学报》2013年第1期166-172,共7页Chinese Journal of Scientific Instrument

基　　金：国家重大专项(2010ZX03006-004);中国科学院知识创新工程重要方向项目(Y022081131)资助

摘　　要：针对机器人快速运动下,由运动模糊而导致视觉里程计定位估计精度下降的问题,结合惯性传感器和视觉里程计提出一种定位算法。该方法以扩展卡尔曼滤波(extended Kalman filter,EKF)为框架,利用惯性传感器的航位推算构建EKF的过程模型,视觉里程计作为相对线速度和相对角速度传感器用来建模观测方程,同时考虑到机器人运动在平面上,在垂直方向和侧向方向不会产生跳动和滑动,利用这两个方向上瞬时速度为零的约束构建另外一个观测方程。提出的定位方法能够克服视觉定位和惯性定位的缺点,提高了定位精度。基于机器人实测数据进行实验,结果表明提出的算法优于单独采用惯性传感器和视觉里程计。Aiming at the problem that the localization estimation accuracy of visual odometry drops due to motion blur under fast robot motion, a localization algorithm is proposed based on inertial sensor and visual odometry. Taking extended Kalmem filter（EKF） as the framework, the proposed algorithm constructs the EKF process model using the dead reckoning of inertial sensor;visual odometry is used as relative linear velocity and angular velocity sensors to model the observation equation. Taking into account the factor that the robot does not slide and jump in vertical and lateral directions while moving on a plane, the constraint that the instantaneous velocities in these two directions are zero is used to construct another observation equation. The proposed localization algorithm overcomes the shortcomings of inertial localization and visual localization and improves localization accuracy as well. Experiments on real robot test data were conducted. Experiment results show that the proposed algorithm is superior to the aoglrithms using inertial sensor or visual odometry alone.

关 键 词：机器人定位 惯性传感器 视觉里程计 扩展卡尔曼滤波 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]