中文摘要：

为了解决水下无人航行器（UUV）携带的测速传感器在失效情况下无法正常导航的问题,提出了一种基于海流剖面的UUV辅助导航方法。该方法首先采用实时性较好的卡尔曼滤波技术对海流剖面数据进行滤波,并引入小波变换方法辅助确定卡尔曼滤波器观测噪声方差、根据导航速度的变化动态计算过程激励噪声方差;然后建立海流剖面-导航速度关系数据库,从而反演UUV的导航速度,进而通过船位推算方法实现对UUV的辅助导航。为验证该方法的有效性,分别就海试时多普勒测速仪（DVL）测得的UUV速度数据1/2有效、1/10有效、1/20有效以及长时无效情况下进行计算分析,试验结果表明：在DVL声呐数据间隔有效的情况下,海流剖面辅助导航可以达到UUV的导航精度（CEP〈0.5%）要求;且即使出现DVL声呐长时间得不到有效数据的情况,其导航精度也大大高于单纯使用UUV动力学模型进行辅助导航时的精度,说明海流剖面辅助UUV导航方法研究具有较大的意义。

英文摘要：

In order to solve the problem of abnormal navigation when the speed-detecting sensor carried in UUV is in- effective, an aided navigation method for UUV based on current profile was proposed. Firstly, Kalman filter technology that has good real time characteristic is applied to filter the ocean current profile data. Wavelet transform is intro- duced to aid Kalman Filter （KF） to observe the noise variance, and the process noise variance is calculated dynami- cally according to the change of navigation speed. Then, the relational database of current profile-navigation speed is established, so the UUV navigation speed can be reversely calculated ; thus the aided navigation for UUV is implemen- ted through dead reckoning. To verify the effectiveness of the aided navigation method ,under the conditions that 1/2, 1/10 and 1/20 of the UUV velocity data measured using a Doppler velocity log （DVL） in sea trial are effective and the measured UUV velocity data are ineffective for a long time, the UUV navigation locations are calculated and ana- lyzed. The results show that when the DVL velocity data are effective in intervals, the aided navigation method based on current profile can meet the requirements of UUV navigation accuracy （CEP 〈0.5% ） ;and even if the DVL sonar could not acquire effective data for a long time, the navigation accuracy is much batter than that obtained by only usingUUV dynamic model method,which indicates that the study of current profile aided navigation method for UUV has significant importance and good application prospect.