基于SR-UKF的纯方位多移动机器人协同定位算法

Bearings-only cooperative localization algorithm of multi-mobile robots based on square-root unscented Kalman filter

为了解决现有多移动机器人定位算法难以同时兼顾定位精度和实时性的问题,该文给出了一种基于相对方位的平方根无迹卡尔曼滤波(Square-root unscented Kalman filter,SR-UKF)多移动机器人协同定位算法。该方法根据机器人运动学方程和量测方程给出多移动机器人自定位的动态模型,利用相对方位作为量测值,在滤波中直接传递协方差矩阵的平方根对系统状态整体更新,实现了多机器人系统的分布式自定位。仿真结果表明:在同等条件下,SR-UKF算法定位精度比已有算法精度提高了近一倍,单次平均运行时间减少了三分之二。

In order to solve the trade-offs between localization accuracy and real-time in the multimobile robots localization,a bearings-only cooperative localization algorithm of multi-mobile robots based on square-root unscented Kalman filter( SR-UKF) is presented. The dynamic model of the multi-mobile robots self-localization system is proposed according to the kinematics equation and the measurement equation. The method uses the relative bearings as measured values,and the square root of covariance matrix is delivered directly in the filtering to update the system state. The distributed self-localization is realized. The simulation results show that,under the same conditions,compared with the existing algorithms,the localization accuracy of the proposed SR-UKF algorithm is increased by nearly one time,and the single average execution time is reduced by 2 /3.