摘要
针对智能船舶依赖单一感知传感器获取的信息难以满足高精度和高可靠性的导航需求,提出基于因子图优化的数据融合方法,该方法结合了惯性导航系统(inertial navigation system,INS)和全球定位系统(global positioning system,GPS)的数据。根据惯性测量单元(inertial measurement unit,IMU)的测量模型对INS进行机械编排。利用预积分原理构建预积分因子节点,利用陀螺仪和加速度计的漂移误差构建漂移因子节点对预积分结果进行修正。通过插入GPS因子,构建INS/GPS因子图组合导航模型,识别并更新受新测量因子影响的部分导航状态变量节点。应用非线性优化等方法获取最优状态估计值,实现多源感知的信息融合。仿真结果验证了该方法的有效性。与常用的卡尔曼滤波相比,在GPS信号故障时,所提出方法具有更好的收敛性能。
For intelligent ships that depend solely on information obtained from a single perception sensor,it is difficult to meet the navigation demand for high precision and high reliability.For the problem,a data fusion method based on factor graph optimization is proposed,and the method combines data from inertial navigation system(INS)and global positioning system(GPS).The mechanical arrangement of INS is carried out according to the measurement model of the inertial measurement unit(IMU).Using pre-integration principle,the pre-integration factor nodes are constructed.The drift errors of gyroscopes and accelerometers are used to construct the drift factor nodes to correct the pre-integration results.An INS/GPS factor graph integrated navigation model is constructed by inserting GPS factors,and some navigation state variable nodes affected by new measurement factors are identified and updated.The optimal state estimation value is obtained by methods such as nonlinear optimization,and the multi-source perception information fusion is realized.The simulation results verify the effectiveness of the method.Compared with the common Kalman filter,the proposed method has better convergence performance when GPS signal fails.
作者
张闯
贾春林
郭沐壮
ZHANG Chuang;JIA Chunlin;GUO Muzhuang(Navigation College,Dalian Maritime University,Dalian 116026,Liaoning,China)
出处
《上海海事大学学报》
北大核心
2024年第4期8-16,共9页
Journal of Shanghai Maritime University
基金
辽宁省应用基础研究计划项目(2022JH2101300265)
中央高校基本科研业务费智库专项课题资助(3132023727)。
关键词
智能船舶
因子图
预积分
组合导航
intelligent ship
factor graph
pre-integration
integrated navigation
作者简介
张闯(1980-),男,辽宁昌图人,副教授,博士,研究方向为船舶组合导航,(E-mail)zhangchuang@dlmu.edu.cn。
