【摘要】 从改进提议分布的成片野值容错能力入手,提出了基于残差正交判别的UPF容错滤波算法.该算法将残差正交判别法UKF的野值自适应性和粒子滤波的"适者生存性"有机地结合起来.通过非线性状态估计的实验,证实了这种新的自适应粒子滤波对成片野值处理的有效性.更多还原

【Abstract】 By improving the fault-tolerance of the proposal distribution consecutive outliers the fault-tolerant filtering algorithm based on residual orthogonal unscented particle filter(UPF) was presented.The algorithm combined the outliers’adaptation of residual orthogonal UPF with the agility of the particle filter perfectly.By experiments of non-linear state estimation it is verified that the new adaptive particle filtering is of great effectiveness to consecutive outliers processing.更多还原