Research On The Arithmetic Of GPS/INS Integrated Navigation System For Long-range Autonomous Underwater Vehicle

Long range and the complex ocean conditions make navigation and position very important for the long-range Autonomous Underwater Vehicle (AUV). To improve the navigation precision, in addition to the high-powered hardwares, high-effective and high-precise filtering arithmetic is effective for restraining the speed of parameter errors. The main work of the dissertation is:1. In the arithmetic optimization of inertial navigation, the key mission is initial extraction and synchro update of the quaternion. The quaternion is extracted from the direction cosine matrix; it's a simple way of optimizing initial extraction of the quaternion. When optimizing synchro update of the quaternion, the quaternion differential equation is calculated by using minimum parametric method. It's proved that the algorithm optimization of quaternion raise the precision and the calculate efficiency.2. Based on the dynamic modeling in state space, the state space model of GPS ise established: constant speed model and constant acceleration model. After analysing the central errors of GPS, constant speed error model of GPS is established.3. As the filter gain and the one-step prediction matrix is improved. The ability and stability of the filter algorithm in tracking vehicle status is raised. The algorithm of matrix inversion with matrix exterior product arithmetic is improved, so that matrix inversion is avoided.4. The SINS/GPS integrated navigation system is designed based on the improved Strapdown filter. The computer simulation is completed through Matlab tools. By comparing classical Kalman filter with improved Kalman filter; it's shown that the improved Kalman filter is better than classical Kalman filter in tracking ability, stability and accuracy.