Research On Track Planning And Image Matching Methods Of Gravity-Aided Inertial Navigation

Inertial navigation systems have the disadvantage of accumulating errors over time.In order to meet the needs of long-term,autonomous,high-precision and covert navigation of underwater vehicles,gravity-aided inertial navigation methods can be used to correct inertial navigation errors.At present,gravity-aided inertial navigation mainly uses a one-dimensional line matching method.Compared with two-dimensional image matching,it uses less effective gravity map information and is prone to problems such as poor matching accuracy and even mismatching in areas where gravity changes are relatively flat.Based on this,the thesis conducts in-depth analysis and research on the key technologies of gravity two-dimensional image matching navigation-selection of adaptation areas,track planning and image matching algorithms.The main research work and innovations are as follows:1.Summarizes the domestic and international research progress of underwater gravity-aided inertial navigation,analyzes and summarizes the key technologies and problems to be solved of underwater gravity-assisted inertial navigation.The basic principles and basic methods for obtaining ocean gravity anomalies such as satellite altimetry,aviation gravity measurement and shipborne gravity measurement are studied,and the research directions and key research contents of the paper are combed.2.Research on the selection method of gravity-aided navigation adaptation area based on multi-eigenvalue analysis.Based on the statistical characteristic parameters of gravity field,the in-depth analysis of the selection area based on information entropy,the selection of adaptive area based on fuzzy comprehensive decision-making,and the Selection method of adaptive region for component analysis.Based on the grid average gravity anomaly numerical model as the basic data,the experimental area is selected and the gravity field statistical characteristic parameters are calculated by partitioning.The multi-index comprehensive selection of the adaptation area is performed,and the experimental calculation analysis is performed by the Iterated Closest Contour Point matching algorithm.The results show that the gravity matching effect of the adaptive region selected by the three methods is better,and the results of the fuzzy comprehensive decision method and the principal component analysis method are better than the information entropy method.3.The technology and method of trajectory planning using ant colony-potential field algorithm are studied.First,based on gravity multi-eigenvalue analysis,the adaptive division of the underwater submarine navigation area is given,and the adaptive and non-adaptive zone labels are given;then the artificial potential field algorithm is introduced based on the ant colony algorithm for track planning,Reconstruct the heuristic function to avoid the local optimal problem of the ant colony algorithm;at the same time,the maximum-minimum ant colony system is used to improve the algorithm pheromone update rules to prevent the "premature" phenomenon.In the experimental environment of the thesis,the efficiency of ant colony-potential field algorithm is improved by more than 40%.Simulation experiment results show that the ant colony-potential field algorithm proposed in this paper can effectively solve the problem of trajectory optimization of underwater submersibles in gravity-aided navigation,and improve the feasibility of the solution.4.A matching algorithm using gravity 2D images to assist inertial navigation is studied.The matching process consists of two parts: coarse matching and fine matching.First,the gravity anomaly data in the area to be matched(2＇×2＇)is grayed to obtain a gravity two-dimensional image measurement template map.Considering the inertial navigation error,the Underwater Vehicle is the center and the confidence radius is 10 nautical miles gravity background field map(20＇×20＇).Secondly,the rough matching process is performed by the gray matching method to obtain the gravity of 5 approximate areas two-dimensional image sub-picture.Finally,feature matching is used to perform the precise matching process,and the final matching result is obtained by comparing the difference between the template image and the sub-picture feature vector.In the experimental environment of the thesis,the grid size of the two-dimensional gravity image is set to 20〃×20〃,and the positioning error is at 1’.Simulation experiment results show that the gravity two-dimensional image-aided navigation algorithm proposed in this paper has a good matching effect and has good practical significance.