【摘要】 为解决“飞点”对滤波效果产生的不利影响,从统计学原理出发,提出一种抗“飞点”的UKF算法.在UKF算法中引入门限函数向量,对新息(innovations)进行控制,既充分利用正常新息确保滤波的精度,又有效抑制异常新息的不利影响,从而达到对“飞点”数据的容错能力.仿真实验验证了抗“飞点”的容错UKF算法的有效性和实用性.更多还原

【Abstract】 To eliminate the unfavorable influence of outliers on the filter,a new unscented Kalman filter(UKF) algorithm for outliers rejection was proposed in terms of statistical principles.The vector of threshold function brought into the algorithm has a control over innovations,which can not only take advantage of normal innovations to increase filtering precision,but also efficiently restrain the unfavorable influence out of outliers.The filter remarkably improves the fault tolerance to outliers.By simulation results it proves that the proposed fault tolerant UKF algorithm is available for outliers rejection.更多还原