基于SE(3)的鲁棒自适应算法及其在SINS/DVL中的应用  

Robust Adaptive Algorithm Based on SE(3)and Its Application in SINS/DVL

作　　者：钱镭源 覃方君[1] 李开龙[1] 朱天高 QIAN Leiyuan;QIN Fangjun;LI Kailong;ZHU Tiangao(School of Electrical Engineering,Naval University of Engineering,Wuhan 430033,China)

机构地区：[1]海军工程大学电气工程学院,武汉430033

出　　处：《上海交通大学学报》2024年第4期498-510,共13页Journal of Shanghai Jiaotong University

基　　金：国家自然科学基金(42274013,61873275)资助项目。

摘　　要：针对复杂环境下SINS/DVL组合导航易受干扰的问题,提出了基于SE(3)的鲁棒自适应算法.通过将李群/李代数理论和鲁棒自适应策略引入正交变换容积卡尔曼滤波(TCKF),使TCKF的估计状态纳入特殊欧氏群,改善了状态空间不一致问题.并利用卡方检验和Huber方法,在量测更新时根据新息向量自适应地重构异常量测.SINS/DVL实验结果表明,所提方法具有比传统方法更优的空间一致性和鲁棒性.Aimed at the problem that SINS/DVL integrated navigation is vulnerable to interference in complex environment,a robust adaptive algorithm based on SE(3)is proposed.By introducing the Lie group/Lie algebra theory and the robust adaptive strategy into the orthogonal transformed cubature Kalman filter(TCKF),the estimated states of TCKF are made to incorporate the Special Euclidean group(SE(3)),and the state space inconsistency problem is improved.The anomaly measurement is adaptively reconstructed based on the innovation vector,by using the chi-square test and the Huber method.The experimental results of SINS/DVL show that the proposed method has a better spatial consistency and robustness than the traditional methods.

关 键 词：组合导航 李群/李代数 Huber方法 正交变换容积卡尔曼滤波 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程] U666.11[自动化与计算机技术—控制科学与工程]