传感器数据异常下动力定位鲁棒状态估计方法

删除或更新信息，请邮件至freekaoyan#163.com(#换成@)

本站小编 Free考研考试/2024-01-16

-->

蒋帆,徐海祥,冯辉,卜瑞波.传感器数据异常下动力定位鲁棒状态估计方法[J].,2020,60(6):610-618

传感器数据异常下动力定位鲁棒状态估计方法

Robust state estimation method for dynamic positioning under abnormal sensor data

DOI：10.7511/dllgxb202006008

中文关键词:动力定位状态估计鲁棒无迹卡尔曼滤波传感器数据异常假设检验

英文关键词:dynamic positioning state estimation robust unscented Kalman filter abnormal sensor data hypothesis test

基金项目:国家自然科学基金资助项目(5187921051979210)；中央高校基本科研业务费专项资金资助项目(2019III0402019III132CG)；武汉理工大学研究生优秀学位论文培育项目(2018-YS-019).

作者	单位
蒋帆,徐海祥,冯辉,卜瑞波

摘要点击次数:252

全文下载次数:217

中文摘要:

针对时变环境带来的传感器数据异常、未知环境干扰等影响，导致基于无迹卡尔曼滤波的动力定位状态估计方法估计精度下降的问题，提出了一种鲁棒无迹卡尔曼滤波方法，该方法通过引入一种基于指数加权的观测噪声协方差矩阵R自适应更新模块和一种基于卡方分布假设检验方法的过程不确定性识别模块处理传感器数据异常情况并估计未知环境力．最后，以某平台供应船的船模为仿真对象，进行了仿真对比实验．仿真结果表明，鲁棒无迹卡尔曼滤波能够准确及时地识别传感器数据异常情况，相较于传统无迹卡尔曼滤波而言，鲁棒无迹卡尔曼滤波状态估计精度更高，收敛速度更快，表现出较强的鲁棒性．

英文摘要:

Aiming at the problem of abnormal sensor data and unknown environmental interference caused by time-varying environment, which causes the estimation accuracy of state estimation method for dynamic positioning based on unscented Kalman filter to decrease, a robust unscented Kalman filter method is proposed. The method adopts an exponentially weighted observation noise covariance matrix R adaptive update module and a process uncertainty recognition module based on the chi-squared distribution hypothesis test method to handle sensor data anomalies and estimate unknown environmental forces. Finally, a simulation experiment is carried out with the ship model of a platform supply ship as the simulation object. Simulation results show that the robust unscented Kalman filter can accurately and timely identify anomalies in sensor data. Compared with the traditional unscented Kalman filter, the robust unscented Kalman filter has higher state estimation accuracy, faster convergence and better robustness.

查看全文查看/发表评论下载PDF阅读器

关闭

传感器数据异常下动力定位鲁棒状态估计方法

本站小编 Free考研考试/2024-01-16

相关话题/

领限时大额优惠券,享本站正版考研考试资料!