摘要
为提高自主式水下潜器(AUV)捷联惯导(SINS)的自主导航精度,增强系统鲁棒性,采用了有关H∞滤波的AUV模型辅助的组合导航方法。克服当海底深度超出多普勒测速仪(DVL)范围时,SINS/DVL组合模式失效导致的捷联惯导误差迅速增大问题。针对传统卡尔曼滤波中系统模型和外部干扰统计噪声不确定时,系统精度降低的情况,提出采用H∞滤波提高系统精度和鲁棒性。在Matlab/LabVIEW的仿真平台上验证组合方法的可行性,仿真结果表明,基于H∞滤波的模型辅助捷联惯导方法能够有效提高组合导航系统精度、鲁棒性,是DVL工作失效时的备份导航系统,并能够有效抑制AUV在模型不准确的情况下Kalman滤波的发散问题。
In order to improve the independent navigation accuracy of autonomous underwater vehicle （AUV） strapdown inertial navigation（ SINS）, and enhance the system robustness, a model -aided integrated navigation for AUV based on H∞ filtering was proposed. When the bottom depth is beyond of the DVL detection region, the SINS/ DVL integrated mode fails. Aiming at the situation that noise statistics and system model of traditional Kalman filte- ring are unknown, Hw filtering was adopted to improve the system precision and robustness. The integrated naviga- tion simulation platform was realized based on Matlab/LabVIEW which can test the feasibility of integrated navigation method. The simulations results show that the model - aided SINS method based on Hw filtering can improve the in- tegrated navigation system precision and robustness effectively. When DVL fails to work, it is a backup navigation system, and can effectively restrain the Kalman filtering divergence with inaccuracy model.
出处
《计算机仿真》
CSCD
北大核心
2014年第2期405-410,共6页
Computer Simulation
基金
国家自然科学基金(61001154
61201409)
中国博士后科学基金(2012M510923)
关键词
组合导航
仿真平台
捷联惯导
模型辅助
Integrated navigation
Simulation platform
Strapdown inertial navigation
Model - aided
