基于改进UKF算法的移动机器人定位方法研究

针对移动机器人在多传感器融合定位过程中因噪声统计特性未知或不准确引起的定位精度不高的问题，提出了一种基于Sage-Husa滤波改进的无损卡尔曼滤波(UKF)移动机器人定位算法。首先建立了移动机器人定位相关模型；然后根据噪声统计特性时变特点利用Sage-Husa中的噪声估计器，对状态噪声和量测噪声进行自适应地估计，减小扰动噪声给定位解算带来的误差；接着在状态更新时引入收敛因子，加快算法收敛速度；最后将UKF算法和改进的UKF算法应用到实验室移动机器人中进行仿真实验。实验结果表明，所提出的算法对状态扰动具有较强的抵制能力，对机器人定位的准确性与稳定性的提升具有显著效果。

Research on mobile robot localization based on improved UKF algorithm

This paper proposed an improved unscented Kalman filter(UKF) mobile robot localization algorithm based on Sage-Husa filter to solve the problem of low localization accuracy caused by unknown or inaccurate noise statistical characteristics in multi-sensor fusion localization process. Firstly, it established the mobile robot localization model. Then, according to the time-varying characteristics of noise statistics, this paper adaptively estimated the state noise and measurement noise to reduce the error of position calculation caused by perturbation noise. The paper also introduced the convergence factor in the state update to accelerate the convergence speed of the algorithm. Finally, it applied the UKF and the improved UKF algorithm to the laboratory mobile robot for simulation experiments. The experimental results show that the proposed algorithm has strong resistance to state disturbance, and has remarkable effect on improving the accuracy and stability of robot positioning.