一种基于分布式EKF的多机器人协同定位方法

A Cooperative Localization Approach for Multiple Robots Based on Distributed Extended Kalman Filter

研究多机器人编队的导航问题,针对多机器人传感器带来的噪声信号影响定位精度,为了提高系统定位精度,提出了一种基于分布式扩展卡尔曼滤波(DEKF)的多机器人协同定位方法。根据航位推算建立单机器人跟踪目标的定位模型后建立编队多机器人的协同定位模型,利用量测信息,通过扩展卡尔曼滤波(EKF)估计各编队机器人局部定位信息,将所得的局部估计值利用优化的融合规则进行处理,得到编队中各机器人的定位数据。通过对单机器人和编队多机器人协同定位进行仿真,结果表明,编队机器人能够利用协同定位方法进行实时定位,且具有更高的定位精度。

In order to enhance the localization accuracy of the muhi-robot formation navigation, a multi-robot co- operative localization method based on distributed extended Kalman filter （DEKF） is proposed. Firstly, the single ro- bot localization model is designed by using the dead-reckoning method, and the cooperative localization model of the formation multi-robot is obtained by employing the measurable information. Secondly, the local localization values of each robot are estimated through the extended Kalman filter （EKF） approach on the measurable information among the robots, and the position of each robot is obtained by fusing the local localization values based on fusion formula. Finally, the comparisons between the new method and the single-robot localization method based on EKF are conduc- ted through a simulation. The simulation results indicate that the formation robots can utilize this coordination locali- zation method for real-time locating, and this method has a higher positioning accuracy than single-robot localization method based on EKF.