Research On The Filter Algorithm Of Deep Sea Long Base Line Navigation System

As an important tool for ocean exploration and development, Human Occupied Vehicle(HOV) has a very wide use both in the field of military, civil and scientific research. The navigation and positioning technology is the critical technology for HOV. With the increasingly deep ocean exploitation, it will put forward higher requirements of accuracy of navigation system. Due to hardware technology and cost constraints, it becomes more and more difficult to get high-precision navigation system through the high-precision sensors. In addition, navigation system with a high degree of autonomy, reliability and anti-interference is essential for the sake of scientific survey and safety in very hazard environment, but navigation system consisting of a single sensor is difficult to meet the requirements. So the high-precision adaptable nonlinear filtering algorithm has become one of the main ways to obtain reliable high-precision navigation system. Unscented Kalman Filter(UKF) is one of the most effective approaches for nonlinear problem, and Square Root Unscented Kalman Filter(SRUKF) has significant advantages compared to the former. With the background of practical scientific research, the dissertation mainly investigates SRUKF and its application in LBL/DR integrated navigation system.Firstly, this paper briefly introduces the system structure and working proceeding of LBL/DR integrated navigation system.Secondly, the application premise and implementation process of UKF and SRUKF are analyzed. Compared with UKF, SRUKF can effectively avoid the filter divergence problem and ca improve the computational efficiency are proved in theory.Thirdly, it concentrates on the drawbacks of SRUKF, such as the shortcoming of measurement update procedure and do not have adaptive ability of dealing with change of noise statistics. The traditional SRUKF is improved to handle these drawbacks through different strategies. Some people apply an iterative measurement updating method easy in realization to SRUKF; the output of the filter estimate with higher precision and smaller variance will be obtain by using the new iterative SRUKF method. In case of the noise statistic being unknown or variable, the SRUKF may lead to the decrease of the estimation precision of the filter or even filter divergence. A new adaptive SRUKF with noise statistical estimator based on maximum a posteriori estimation and reducing memory index weighting is proposed. In the process of filtering, for one thing it can use the measured value correct predicted value, for another it can estimate of the unknown or time-varying noise statistical characteristics.Finally, data obtained by HOV named Jiaolong in previous sea trial is used to verify the validity and rationality of iterative SRUKF and adaptive SRUKF.