Research On Key Technologies Of GNSS-A Underwater Positioning And Navigation

Marine positioning and navigation are the foundation of marine scientific research,maritime transportation,marine rights maintaince,marine resource development,marine engineering construction,marine environmental governance,marine battlefield construction,which provides whole-directional,whole-process,whole-time,multi-space,multi-level,multi-link marine spatiotemporal information and location services for all human maritime activities.With the deepening of marine development activities,the requirements for marine positioning and navigation have changed from above the water surface to all marine space.Although the GNSS(Global Navigation Satellite System)has greatly promoted the new development in the field of geodetic surveying,the rapid attenuation of electromagnetic waves in seawater makes it impossible to directly use for underwater target positioning and navigation.Due to the good propagation characteristics of sound waves in seawater,GNSS-A(Global Navigation Satellite System-Acoustic)positioning and navigation technology has been widely studied and applied.However,complex dynamic marine environment changes have brought a series of key technical problems to the high-precision underwater positioning and navigation.This paper systematically summarized the reaearch status of underwater positioning and navigation,focusing on the problems of sound ray tracing,difference underwater positioning and underwater navigation filtering,carried out systematic analysis,research and improvement.The main contents are as follows:(1)Sound ray tracing method of underwater positioning considering incident beam angleThe spatial and temporal variation of sound velocity in seawater can make sound waves refract along the propagation direction,and effective elimination of refraction artifacts is very important to improve the accuracy of underwater acoustic positioning.When the sound velocity profile is known,sound ray tracing is an effective method to reduce the refraction artifacts.The existing sound ray tracing methods require that the incident beam angle is known,while the underwater acoustic positioning systems based on the distance intersection principle usually do not directly measure the incident beam angle.A sound ray tracing method of underwater positioning considering incident beam angle is proposed,which can trace sound ray and estimate target position respectively by iterative calculation,and the incident beam angle is determined by a search method.In order to improve the calculation efficiency,a transcendental equation solution method is proposed to determine the incident beam angle iteratively.The experiment shows that the proposed method can effectively eliminate the effect of the refraction artifacts by using sound velocity profile,and the calculation efficiency of the transcendental equation solution method is better than the search method.(2)Static single difference positioning between epochs and its ill-posed solutionThe static single difference underwater positioning method between epochs eliminates the influence of partial systematic error by making difference between adjacent observation epoch,which has not been widely applied due to the possible ill-posed peoblem.By analyzing the ill-posed degree of single difference positioning method between epochs under different measurement conditions(track,heave),it is found that the difference of geometric incident angle is one of the main factors of ill-posed problem,then the regular polygon tracks are used to reduce the ill-posed degree.In order to solve the ill-posed problem of single difference positioning between epochs,a LIU-type estimation method based on improved L curve is proposed,which using L curve composed of mean square error and residual norm to optimize the parameters of LIU-type estimation.The experiment shows that the static single difference underwater positioning method between epochs can eliminate the influence of systematic error,and the improved LIU-type estimation method can effectively improve the estimation results of ill-posed problem.(3)Dynamic single difference positioning between reference points and its network designAccording to the problem that underwater dynamic non-difference positioning method using multi-reference points cannot eliminate the influence of systematic errors,a dynamic single difference underwater positioning method between reference points is proposed.By calculating the difference between the distance observations of the same observation epoch,the influence of partial systematic errors is eliminated.The network structure of dynamic single difference positioning method between reference points is designed and optimized based on the Position Dilution of Precision(PDOP).Regarding the regular polygon network with the vertex difference reference point,because the direction cosine vectors are approximately equal,the network structure is poor.Therefore,a regular polygon network with the central difference reference point is proposed for network optimization.The experiments of LBL underwater positioning show that the single difference positioning method between reference points can effectively eliminate the influence of systematic errors,and the regular polygon network with the central difference reference point can enhance the strength of the network structure.(4)Underwater navigation Kalman filter algorithm based on Interacting Multiple ModelBuilding a motion model correspond to the actual montion state of the underwater target is an important premise to ensure accuracy and reliability of underwater navigation Kalman filter.However,in the influence of complex marine environment,the motion state of underwater dynamic target has strong diversity and randomness,so it is impossible to adopt the motion model exactly matching the actual motion state of the target according to prior information.When the motion state of the underwater target is switched between maneuvering and non-maneuvering modes,the single maneuvering or non-maneuvering motion model will make the filtering accuracy decrease or even the filtering diverges.Therefore,the underwater navigation Kalman filter algorthim based on interacting multiple model is proposed,and the weighted fusion of multiple models is realized by using Markov prior transfer probability.The simulation experiment show that compared with the Kalman filter algorithm based on single model the Kalman filter algorithm based on Interacting Multiple Model has the better adaptability of montion state.