一种稳健的自适应传递对准算法  

Robust Adaptive Algorithm for Transfer Alignment

作　　者：徐林[1] 李世玲[1] 屈新芬[1] 

机构地区：[1]中国工程物理研究院电子工程研究所,绵阳621900

出　　处：《系统仿真学报》2012年第11期2324-2328,共5页Journal of System Simulation

基　　金：中国工程物理研究院电子工程研究所创新基金(S20090101)

摘　　要：传递对准算法的性能受量测噪声影响较大但其统计特性通常难以精确给定。为实现对传递对准中量测噪声的稳健自适应调整,提出了一种改进的自适应传递对准算法。在传统的Sage-Husa自适应卡尔曼滤波算法基础上,通过减少自适应调整的参数个数、采用渐进无偏估计代替无偏估计、分段调整加权系数三重处理,有效地提高了算法的稳健性,同时算法的计算量也有所降低。结合一种经典的速度加姿态匹配传递对准模型,对改进的自适应算法进行了仿真测试。仿真结果表明,当量测噪声协方差阵取值与真实噪声水平偏离时,该改进算法能够对量测噪声的统计特性进行有效的自适应调整,从而使传递对准在精度和稳健性两个方面的综合性能得以改善。Performance of transfer alignment algorithm greatly relies on the prior statistical property of measurement noise, which is usually difficult to determine. To solve this problem, an improved transfer alignment algorithm aiming at robust and adaptive adjustment of the statistical property was investigated. Three modifications, namely reducing parameters to be adjusted, using asymptotically unbiased noise estimator and piece-wise weighted coefficients, were carried out on traditional Sage-Husa adaptive Kalman filtering algorithm, and consequently a simplified but more robust adaptive filtering algorithm was obtained. With a classical transfer alignment model of velocity plus attitude matching, the efficiency of this new algorithm was checked by simulation. It is indicated that by using this algorithm, the statistical property of measurement noise can be adjusted adaptively if it deviates from its true value, and thus both the accuracy and robustness of transfer alignment are improved.

关 键 词：传递对准 自适应卡尔曼滤波 捷联惯导系统 速度加姿态匹配 稳健性 

分 类 号：V249.32[航空宇航科学与技术—飞行器设计]