本文为美国海军研究生院（作者：David W. McClarin）的硕士论文，共140页。

凤凰号自主水下航行器必须能够始终准确地确定其位置，这就要求：1）用于水面导航的GPS和差分GPS； 2）用于水下导航的短基线声纳测距系统；以及3）位置的数学模型。本文介绍了一种卡尔曼滤波方法，用于合并GPS、差分GPS、短基线声纳测距和数学模型，以生成航行器位置和洋流的单个状态向量。滤波器以扩展模式运行以处理非线性声纳范围，并以正常模式运行以处理线性GPS DGPS数据。这需要安装GPS系统，并确定这些系统之间的差异和误差。现在，凤凰号已经具有使用位置测量系统的实时位置确定方法，可以单独使用或组合使用。这项工作的结果已通过在海上对航行器进行的实际测试得到了验证，在该测试中获得了精确到几米的位置估计。

The Phoenix Autonomous Underwater Vehiclemust be able to accurately determine its position at all times. Thls requires:1) GPS and differential GPS for surface navigation, 2) short baseline sonarranging system for submerged navigation, and 3) mathematical modeling ofposition. This thesis describes a method of Kalman filtering to merge the GPS,differential GPS, short baseline sonar ranging, and the mathematical model toproduce a single state vector of vehicle position and ocean currents. Thefilter operates in the extended mode for processing the non-linear sonarranges, and in normal mode for the linear GPSDGPS data. This required installationof a GPS system and the determination of the different variances and errorsbetween these systems. Phoenix now has a real time method of positiondetermination using either position measuring system separately or combined.The results of this work have been validated by real world testing of thevehicle at sea, where position estimates accurate to within several meters wereobtained.

1 引言

2 凤凰号硬件特点及其不足

3 卡尔曼滤波

4 导航

5 软件

6 小结与结论

附录A NAVIGATOR1.C

附录B KALMAN_FILTER.H

附录C KALMAN_FILTER.C

附录D READGPS.H

附录E READGPS.C

附录F MATRIX.H

附录G MATRIX.C

更多精彩文章请关注公众号：