基于自适应渐消EKF的FastSLAM算法

快速同时定位与建图(fast simultaneous localization and mapping, FastSLAM)算法的采样过程会带来粒子退化问题,为了改进算法的性能,提高估计精度,从研究粒子滤波的建议分布函数出发,提出基于自适应渐消扩展卡尔曼滤波(adaptive fading extended Kalman filter, AFEKF)的FastSLAM算法。该算法基于FastSLAM的基本框架,利用AFEKF产生一种参数可自适应调节的建议分布函数,使其更接近移动机器人的后验位姿概率分布,减缓粒子集的退化。因此在同等粒子数的情况下,该算法有效提高了SLAM精度,以此减少所使用的粒子数,降低算法的复杂度。基于模拟器和标准数据集的实验仿真结果验证了该算法的有效性。

FastSLAM algorithm based on adaptive fading extended Kalman filter

Sampling process often causes particle degradation in fast simultaneous localization and mapping （FastSLAM）. From the point view of the proposal distribution function, a method named the FastSLAM based on adaptive fading extended Kalman filter is proposed to improve the performance of the algorithm and increase estimation accuracy. It uses the adaptive fading extended Kalman filter (AFEKF) to compute proposal distribution based on the basic framework of FastSLAM, then this proposal distribution is more close to the posterior position of the mobile robot and the degree of particle degradation is reduced. In the case of the same number of particles, the algorithm can effectively improve the accuracy of SLAM. Hence it can reduce the number of particles used in the algorithm and the complexity of the algorithm. The validity of the proposed algorithm is verified by the experimental simulation results based on the simulator and the standard data set.