论文部分内容阅读
针对标准Fast SLAM算法存在的雅可比矩阵的计算、线性化误差累积等问题,提出了一种球面单径容积Fast SLAM算法(SSRCFast SLAM).算法的特点在于使用3阶球面单径准则计算SLAM中的非线性高斯权重积分,以提高精度.所提算法利用球面单径容积粒子滤波进行路径估计,利用球面单径容积卡尔曼滤波来维护路标.通过仿真实验和维多利亚公园数据集实验将所提算法同Fast SLAM2.0、UFast SLAM和CFast SLAM进行对比.结果显示,所提算法在不同粒子数与噪声环境下的定位与建图能力均优于其他3种算法,且在粒子数目较少或环境干扰较大时优势更显著,验证了所提算法的优越性.
Aiming at the problems of standard Fast SLAM algorithm, such as Jacobian matrix calculation and linearization error accumulation, a new algorithm called SSRCFast SLAM is proposed. The algorithm is characterized by using SLM In order to improve the accuracy.The proposed algorithm uses spherical single-diameter volumetric particle filter to estimate the path and uses spherical single-diameter volume Kalman filter to maintain the road mark.According to the simulation experiment and Victoria Park dataset experiment, the proposed algorithm Compared with Fast SLAM2.0, UFast SLAM and CFast SLAM, the results show that the proposed algorithm is better than the other three algorithms in positioning and mapping ability under different particle numbers and noises, The advantage is more significant when the interference is larger, which verifies the superiority of the proposed algorithm.