基于回朔式EKF的多传感器融合在移动机器人自定位中的应用

Application of retroactive EKF-based multisensor fusion to the self-localization of mobile robot

提出了一种基于扩展卡尔曼滤波器 (EKF)的多传感器融合系统 ,用于融合多个超声波传感器及光电编码器的测量值 ,并以该融合值来复位光电编码器 ,克服其累积误差的影响 .由于扩展卡尔曼滤波器是一项耗时的计算任务 ,为了满足实时性的要求 ,采用一种基于回朔式的算法 .

Fusing the multisensor datum can significantly raise the precision of the self-localization. In this paper, a kind of multisensor fusion system based on Extended Kalman Filter(EKF) is presented which is used to fuse the measurements of multiple ultrasonic sensors and photoelectric code-recorder. And photoelectric code-recorder is then reset by the fusion result in order to eliminate the influence of its cumulative errors. Because EKF is a time-consuming calculation, a kind of retroactive algorithm is adopted in order to satisfy the real-time requirement. Simulation result shows the validity of this algorithm.