摘要：This paper presents a robust adaptive interpolation particle filtering algorithm, which overcomes the drawbacks of particle filtering, such as the particle would produce degradation, and the importance density function is difficult to choose. The algorithm absorbs the advantages of robust adaptive filtering, second-order interpolation filtering and particle filtering. It obtains the importance density function by second-order interpolation filtering, and controls the effects of dynamic model errors and abnormal observation on the navigation solution by robust adaptive factor. The proposed algorithm is applied to SINS/CNS/SAR integrated navigation system, and compared with the classical particle filtering. The simulation results demonstrate that the attitude error, velocity error and position error of the robust adaptive interpolation particle filtering are within [-0.3',+0.3'], [-0.4 m/s,+0.4 m/s] and [-5 m,+5 m] respectively, and the filtering performance is significantly superior to that of the classical particle filtering. The new filtering algorithm can not only effectively inhibit the particle degradation, but also effectively control the impact of dynamic model error and abnormal observation, thus improving the accuracy of integrated navigation filter.%针对粒子滤波存在的粒子退化和重要性密度函数难以选取的问题,在吸收抗差自适应滤波、二阶插值滤波和粒子滤波算法优点的基础上,提出了一种新的抗差自适应插值粒子滤波算法.该算法利用二阶插值滤波算法得到重要性密度函数,通过抗差自适应因子实时控制动力学模型误差及观测异常对导航解的影响.将该算法应用于SINS/CNS/SAR组合导航系统进行计算仿真,并与经典的粒子滤波算法进行比较分析.结果表明,提出的滤波算法得到的姿态误差控制在[-0.3′,+0.3′],速度误差控制在[-0.4 m/s,+0.4 m/s],位置误差控制在[-5 m,+5 m],性能明显优于经典的粒子滤波算法.新的滤波算法不但能够有效地抑制粒子退化,而且能够有效地控制动力学模型误差及观测异常的影响,提高了组合导航的滤波精度.
