Research On Underwater Terrain Matching Navigation Algorithm Based On Gradient Fitting

With the increase of military demand and ocean exploration demand,Autonomous Underwater Vehicle(AUV)has been applied more and more widely in the Marine field.Inertial Navigation System(INS)for the AUV provides the position and posture information,but its error inevitably accumulates over time.The underwater terrain can provide additional information to assist in correcting the cumulative errors of INS.This underwater terrain assisted inertial navigation is a reliable method to estimate the exact position and attitude of AUV.With the support of the key project of national natural science foundation of China,this paper seeks solutions to the specific problems of underwater terrain assisted navigation based on the currently popular multi-beam sounding sonar and inertial integrated navigation system.The main contents and research work of this paper include:(1)Two basic algorithms of underwater terrain aided navigation are introduced: correlation matching algorithm and continuous recursive track tracking algorithm.This paper focuses on the differences between the two algorithms in data processing and the use of underwater digital terrain map,and analyzes the characteristics and applicable conditions of the two algorithms.(2)It introduces the process of underwater topographic survey mapping,studies the gradient characteristics and distribution fitting of real terrain data,and analyzes the effects of different mathematical distributions from the perspective of fitting degree and stability.(3)The corresponding relationship between the selected proportion of terrain data and the gradient threshold is studied,so as to use prior statistical parameters to divide terrain data of different gradient ranges in specific navigation and positioning algorithms.(4)Aiming at the demand of AUV discontinuous and high-precision positioning,a "first rough then fine" correlation matching positioning algorithm is proposed in combination with gradient fitting research.Compared with the two traditional matching and positioning methods,the calculation and positioning accuracy of the proposed algorithm are analyzed.(5)According to the theory of particle filter,an AUV path tracking algorithm based on gradient fitting is proposed for the continuous and longtime navigation of AUV.According to the difference of terrain relief,the superiority of the improved algorithm is analyzed from three aspects: mean square error,algorithm stability and calculation amount.This paper proposes two different topographical assistant navigation schemes,related matching positioning and continuous track tracking,based on the different navigation demands of AUV for maneuverability,timeliness and accuracy,combined with the gradient characteristics and fitting research of underwater terrain.Through off-line simulation data processing,the performance advantages of the proposed algorithms are verified from the influence of terrain relief,mean correction error,algorithm stability and CPU running time,etc.The results show that the two schemes proposed in this paper can meet different AUV navigation and positioning requirements,and have higher reliability and stability compared with the traditional algorithms.