基于迭代无迹卡尔曼滤波的水下组合导航

Underwater Integrated Navigation Based on IUKF

研究了超短基线定位系统、多普勒测速仪以及电子罗盘相结合的水下机器人组合导航问题,提出了一种基于迭代无迹卡尔曼滤波(IUKF)的水下多传感器组合导航融合算法,降低了传统无迹卡尔曼滤波器中过程噪声协方差矩阵和测量噪声协方差矩阵对滤波精度和响应时间的影响。该方法通过对多传感器数据融合结果进行多次迭代,提高了水下多传感器组合导航精度。通过多传感器的水池实验和仿真分析,验证了水下组合导航算法的可行性和有效性。

This paper considers the integrated navigation problem of underwater vehicle based on the combination of ultra-short baseline positioning system,Doppler velocimeter and electronic compass.A fusion algorithm based on iterative unscented Kalman filtering is proposed for multi-sensor underwater integrated navigation.This proposed algorithm can reduce the influence of process noise covariance matrix and measurement noise covariance matrix in traditional unscented Kalman filter on the filtering precision and response time.By iterating the results of multi-sensor data fusion,the proposed method improves the precision of underwater multi-sensor combined navigation.Firstly,the obtained data from the sensors are preprocessed,and the motion trajectory diagram simulated by the ultra-short baseline positioning system,the motion trajectory diagram calculated by dead reckoning and the motion trajectory diagram fused by the unscented Kalman filtering algorithm are drawn,respectively.Secondly,the unscented Kalman filtering algorithm iterates the output of multi-sensor data fusion for achieving better fusion effect.Finally,the combined underwater navigation algorithm is verified via multi-sensor pool test and simulation analysis.By comparing the simulated ultra-short baseline positioning data,the dead reckoning data based on the electronic compass and Doppler velocimeter and the position information of the positioning data fused with the iterative unscented Kalman filtering data,it is shown that the iterative unscented Kalman filtering algorithm based on underwater integrated navigation system can achieve the better performance.