复杂水下环境下高精度SINS/USBL误差标定技术

针对SINS/USBL组合系统在导航与定位前需要对一体化样机进行精确标定的问题,对复杂水下声场环境下高精度SINS/USBL误差标定技术展开了研究,包括对安装误差角和空间杆臂误差的估计.基于标定模型的几何关系,推导了标定的状态方程和量测方程,并提出了新型的基于相对量测信息滤波估计的误差标定技术.为解决标定过程中由于存在声学野值而导致滤波估计性能下降的问题,通过高斯牛顿迭代将Huber M估计嵌入到变分贝叶斯框架中,推导了基于M估计的非线性鲁棒自适应标定算法,在精确校正后无需重复标定.仿真的结果验证了该方案的有效性、实用性和鲁棒性.

High-precision SINS/USBL Error Calibration Technology in Complex Underwater Environment

High-precision SINS/USBL error calibration of the integrated prototype is needed before the navigation and positioning of the SINS/USBL integrated system, and the error calibration technology under the complex underwater sound field environment is studied, including the estimation of the installation angle error and the spatial lever arm error. Based on the geometric relationship of the calibration model, the calibrated state equation and measurement equation are deduced, and a new error calibration technique based on filtering estimation of relative measurement information is proposed. In order to solve the problem of degraded filter estimation performance caused by the presence of acoustic outliers in the calibration process, the Huber M estimation is embedded into the variational Bayesian framework using Gauss-Newton iteration, and a nonlinear robust adaptive calibration algorithm based on M estimation is derived, avoiding repeated calibration once accurately calibrated. The results of simulation and Yangtze River test both verify the effectiveness, practicability and robustness of the proposed strategy.