一种小型无人机的FastSLAM算法

A FastSLAM algorithm for small unmanned aerial vehicle

为解决未知环境中小型无人机自主飞行问题,将SLAM算法从地面机器人的二维环境扩展到小型无人机的三维环境.首先,建立了小型无人机的SLAM算法数学模型,得到其非线性状态方程;然后,将小型无人机的SLAM问题分解成路径估计与环境地标估计两部分,分别采用粒子滤波器和扩展卡尔曼滤波器进行估计,提出了一种小型无人机的FastSLAM算法;最后,分别采用扩展卡尔曼滤波和FastSLAM算法进行仿真实验,仿真结果表明FastSLAM算法具有更好的定位精度.

In order to solve the problem of autonomous flight of the small unmanned aerial vehicle(SUAV)in unknown environment,simultaneous localization and mapping(SLAM)algorithm was expanded from two-dimensional environment of the ground robot to three-dimensional environment of SUAV.First,the mathematic model of SUAV SLAM algorithm was built to obtain the nonlinear state equation of SUAV.Second,the SLAM problem of SUAV was decomposed into the estimation over path using aparticle filter and the estimations over landmarks using extended Kalman filters for the purpose of designing a FastSLAM algorithm for SUAV.Finally,simulation researches respectively based on extended Kalman filter(EKF)and FastSLAM algorithm were carried out on SUAV.Results show that FastSLAM algorithm has better performance in location accuracy than EKF algorithm.