Research On Key Technology Of Underwater Gravity Gradient Assisted Inertial Navigation

In this paper,the key techniques of underwater gravity gradient assisted inertial navigation are studied.Firstly,the basic principle and error compensation method of gravity gradient measurement are analyzed.Then marine gravity gradient background model are established combined with satellite altimetry gravity anomaly and EIGEN-6C4 gravity field model.Finally,the gravity gradient matching navigation algorithm based on different matching quantities is studied emphatically.The main work and innovation of the paper are as follows:1.The main key technologies of gravity gradient assisted inertial navigation are summarized,and the research status and progress of gravity gradient assisted navigation are introduced.2.The measurement principle of the gravity gradient gradient of the two-dimensional rotary accelerometer and the three-dimensional space accelerometer is studied.The error source of the accelerometer-type gravity gradient is analyzed,and the adjustment of the accelerometer scale factor and the gradient of gravity Error compensation analysis.3.The basic method of constructing gravitational gradient tensor based on terrain information is discussed.The optimization calculation of the mathematical model of gravity gradient tensor using EIGEN-6C4 gravity field model is introduced.Finally,combined with satellite altimetry gravity anomaly and EIGEN-6C4 gravity field model,the background model of marine gravitational gradient is established based on the removal-recovery technique.At the same time,the variation of different gradient tensor in the construction area is analyzed,and the adaptability characteristics of different component reference maps are analyzed statistically.The results show that the full tensor reference graph is the most violent and has the richest information and the best matching feature.4.The relative extreme value matching algorithm based on the combination of single gravity gradient tensor and different gravity gradient tensor as matching quantity is studied.The precision comparison and analysis are carried out by matching navigation experiment in a certain marine area.The results show that the T_zz tensor in the single component matching navigation has the best matching result.And the matching accuracy is better than two gravitational grid.For the matching navigation algorithm with different gravity tensor combinations as the matching amount,the correlation mathematical model e based on the new criterion is established,The results show that the matching accuracy based on the combination of gravity gradient full tensor is the best in the four combinations,and it is better than one gravitational grid in the lontitude and latitude directions.The results show that the matching accuracy of the gradient tensor combination matching navigation is more accurate and easy to implement.In order to improve the stability and robustness of gravity gradient tensor combination matching navigation,a full probability weighted correlation matching algorithm is proposed.The results show that the matching result has higher and more stable matching accuracy than the correlation extremum matching algorithm.5.In order to meet the real-time and maneuverability requirements of underwater navigation,the principle of unscented Kalman filter?UKF?and the UT transformation method of symmetric distribution sampling are studied.And experimens were conducted in certain area by using a single gravity gradient tensor and four gravity gradient tensors.The results show that the UKF matching navigation accuracy of single gravity gradient tensor is better than that of other single gravity gradient tensor,and the total precision reaches about three gravitational grid.Combined tensor UKF matching accuracy is superior to a single gravity gradient tensor,and the gravity gradient tensor combination in the four combinations has the best matching results,the total accuracy is better than 2.5 gravitational grid.At the same time,gravity gradient tensor combination matching experiments with different initial errors are calculated and analyzed.The experimental results show that when the initial error is large,the full tensor gravity gradient navigation UKF algorithm is more accurate than the other gravity gradient tensor combination.