基于类电磁机制优化的FastSLAM2.0算法

A FastSLAM2.0 algorithm based on electromagnetism-like mechanism

中文关键词  类电磁机制  FastSLAM  粒子滤波  提议分布  重采样

英文关键词  electromagnetism-like mechanism  FastSLAM  particle filter  proposal distribution  resampling

中文摘要

      针对FastSLAM2.0算法粒子退化和粒子耗尽的问题, 提出一种基于类电磁机制优化的FastSLAM2.0算法. 该算法用无迹卡尔曼滤波(unscented Kalman filter, UKF)替代拓展卡尔曼滤波(extended Kalman filter, EKF)估计后验位姿提议分布减少模型线性化误差, 提高采样粒子的质量; 在重采样中模拟电磁场中带电粒子之间吸引排斥机制, 把采样粒子看成带一定电荷量的电子, 通过类电磁吸引力驱动粒子集朝高似然区域移动, 使之较快分布在机器人真实位姿附近, 缓解粒子退化问题, 同时, 通过类电磁排斥力驱使粒子在移动过程中保持一定距离, 保证了粒子多样性. 仿真结果表明了该算法的有效性.

英文摘要

      To solve the problem of particle degradation and sample dilution in FastSLAM2.0 algorithm, this paper proposed a new FastSLAM2.0 algorithm based on electromagnetism-like mechanism (EM–FastSLAM2.0). Firstly, in order to reduce the model linearization error, unscented Kalman filter (UKF) is used to replace extended Kalman filter (EKF) for posteriori estimation of robot pose in EM–FastSLAM2.0 and the quality of sampling particles are improved. Secondly, assuming that each sampling particle is looked as an charged electron and attraction-repulsion mechanism in electromagnetism field is used to simulate interactive force between the particles, electromagnetism-like mechanism is introduced to resampling procedure, where electromagnetism-like attractions drive the particles approach to high likelihood region and make them distributed closely to the true pose value rapidly which can reduce particle degradation. Simultaneously, electromagnetic-like repulsions drive the particles separate from each other and maintain a certain distance in moving process to assure particle diversity. Finally, the simulation results show the effectiveness of the new algorithm.