【作者】 范颖； 张勇刚；

【Author】 ZHANG Yonggang;FAN Ying;College of Automatic, Harbin Engineering University;

【机构】 哈尔滨工程大学自动化学院；

【摘要】 在自主水下航行器(Automatic Underwater Vehicles,AUVs)的协同定位中,受到洋流的影响,量测噪声会出现野值的情况,此时,基于传统的非线性滤波的AUV协同导航方法不再适用。对此,本文基于最大互相关熵准则,提出了一种最大互相关熵容积卡尔曼滤波(Maximum Correntropy Cubature Kalman Filter,MCCKF),然后设计出了基于MCCKF的AUV协同导航方法。为了选取合适的核宽度σ以获取更好的结果,本文选取了三种不同的核宽度进行湖上试验,得出在σ=20时滤波效果最佳。最后,选取了基于容积卡尔曼滤波(Cubature Kalman Filter,CKF)和Huber无迹卡尔曼滤波(Huber Unscented Kalman Filter,HRUKF)的AUV协同导航方法与本文提出的方法进行湖上试验。实验结果表明,本文设计的基于MCCKF的AUV协同导航方法精度最高。更多还原

【Abstract】 In cooperative localization of autonomous underwater vehicles(AUVs), the measurement noise outliers will appear, which is influenced by ocean current. In this case, traditional cooperative navigation method based on nonlinear filter is no longer applicable. Therefore, based on the maximum correntropy criterion, a novel maximum correntropy cubature Kalman filter(MCCKF) is proposed in this paper. Then the AUV cooperative navigation method based on MCCKF is designed. In order to select an appropriate kernel bandwidth σ, three different kernel bandwidth of the proposed method are texted and the filtering accuracy is best when σ=20. Finally, this paper select the cooperative navigation method based on CKF and HRUKF and the method proposed in this paper to carry out the lake trial. The experimental results show the proposed cooperative navigation method based on MCCKF are convergent. The cooperative location accuracy of the cooperative navigation method proposed in this paper has the best estimation accuracy.更多还原