路飞, 田国会, 刘国良, 王宇恒. 智能空间下基于WIFI指纹定位与粒子滤波的服务机器人复合全局定位系统设计[J]. 机器人, 2016, 38(2): 178-184. DOI: 10.13973/j.cnki.robot.2016.0178
引用本文: 路飞, 田国会, 刘国良, 王宇恒. 智能空间下基于WIFI指纹定位与粒子滤波的服务机器人复合全局定位系统设计[J]. 机器人, 2016, 38(2): 178-184. DOI: 10.13973/j.cnki.robot.2016.0178
LU Fei, TIAN Guohui, LIU Guoliang, WANG Yuheng. A Composed Global Localization System for Service Robot in Intelligent Space Based on Particle Filter Algorithm and WIFI Fingerprint Localization[J]. ROBOT, 2016, 38(2): 178-184. DOI: 10.13973/j.cnki.robot.2016.0178
Citation: LU Fei, TIAN Guohui, LIU Guoliang, WANG Yuheng. A Composed Global Localization System for Service Robot in Intelligent Space Based on Particle Filter Algorithm and WIFI Fingerprint Localization[J]. ROBOT, 2016, 38(2): 178-184. DOI: 10.13973/j.cnki.robot.2016.0178

智能空间下基于WIFI指纹定位与粒子滤波的服务机器人复合全局定位系统设计

A Composed Global Localization System for Service Robot in Intelligent Space Based on Particle Filter Algorithm and WIFI Fingerprint Localization

  • 摘要: 针对机器人在没有任何初始位姿先验知识的情况下,通过传感器感知信息确定位姿的全局定位问题,将智能空间技术与 ROS(robot operating system)服务机器人相结合,设计了一种智能空间技术支持下的基于 WIFI 指纹定位和蒙特卡洛粒子滤波定位的复合服务机器人全局定位系统.在该复合定位方法中,首先利用智能空间中的基于 BP(backpropagation)神经网络的 WIFI 指纹定位对机器人进行粗定位,并将估计位置与估计误差发送给 ROS 服务机器人;在粗定位的基础上使用蒙特卡洛粒子滤波算法方法最终获得服务机器人的精确位置.实验结果表明,本文设计的系统实现了 ROS 机器人与智能空间之间的零配置与松耦合,可有效地提高服务机器人全局定位精度,缩短计算迭代时间.

     

    Abstract: With the combination of intelligent space technology and ROS (robot operating system) service robot, a composed global localization system for service robot based on Monte Carlo particle filter algorithm and WIFI fingerprint localization is presented. This system can solve the global localization problem in which the initial pose of service robot is unknown and the robot pose is determined according to sensor information. In the composed localization method, robot rough localization is realized firstly using the WIFI fingerprint localization based on BP (backpropagation) neural network under the support of intelligent space technology, and the estimated position and estimation error are sent to the ROS service robot. Based on the rough localization result, Monte Carlo particle filter algorithm is adopted to get precise position of the ROS robot. The experiment results show that the proposed system can realize zero-configuration and loose coupling between ROS service robot and intelligent space, improve precision of global localization of the ROS service robot effectively, and reduce the iteration time.

     

/

返回文章
返回