基于SAR图像匹配结果可信度评价的INS/SAR自适应Kalman滤波算法

在惯性导航系统（INS）/合成孔径雷达（SAR）组合导航系统中，SAR图像易受斑点噪声的影响，图像匹配的精度对整个导航系统精度的影响十分明显，能够准确地分析SAR图像匹配过程中的误差特性，利用有效的图像匹配信息辅助INS进行组合定位尤为重要。针对上述问题，在加权Hausdorff距离匹配算法的基础上，对影响SAR图像匹配精度的因素进行了分析，提出了一种基于模糊推理的匹配结果可信度评价准则，经过可信度筛选，将有效的匹配信息与INS进行组合；对合理范围内的匹配误差变化引起量测噪声统计特性发生变化，进而导致Kalman滤波精度下降的问题，研究采用改进的Sage-Husa自适应滤波算法对量测噪声方差阵进行动态调整，使其更加接近系统的当前状态。搭建仿真验证平台对所提算法进行了验证，结果表明，该算法能够在合理的匹配误差范围内，有效地筛选出可信的图像匹配结果，相比常规Kalman滤波算法，显著地提升了INS/SAR组合导航系统水平方向的定位精度。

INS/SAR adaptive Kalman filtering algorithm based on credibility evaluation of SAR image matching results

In Inertial Navigation System (INS)/Synthetic Aperture Radar (SAR) integrated navigation system, SAR image is susceptible to speckle noise. The accuracy of image matching has significant impact on the accuracy of the whole navigation system. It is especially important to accurately analyze the error characteristics in the SAR image matching process and to exploit the effective image matching information to assist the INS for integrated positioning. In response to the problems above, based on the weighted Hausdorff distance matching algorithm, the factors affecting the matching accuracy of SAR images are analyzed. Meanwhile, a set of criteria for credibility evaluation of matching results based on fuzzy reasoning is proposed, integrating the effective matching information with INS after credibility screening. Moreover, the variation of the matching error within a reasonable range will cause the statistical characteristics of the measurement noise to change, leading to the degradation of Kalman filtering accuracy. To address this problem, the improved Sage-Husa adaptive filtering method is adopted to dynamically adjust the measurement noise variance matrix closer to the current state of the system. The simulation verification platform is established to verify the proposed algorithm. The results indicate that the algorithm can effectively screen out the reliable image matching results within a reasonable range of matching error. Compared with the conventional Kalman filtering algorithm, the positioning accuracy of the INS/SAR integrated navigation system in the horizontal direction is significantly improved.