| As a significant cornerstone in the field of oceanographic research,deep Human Occupied Vehicle(HOV)is an indispensable carrying equipment for scientific research and survey operations in the deep sea.In the complex and changeable marine environment,high-precision navigation and positioning technology is one of the key and bottleneck technologies for HOV to accurately reach the target point,operate and return safely.The objects of this research are the key technologies of integrated navigation system for deep HOV,mainly including nearseabed integrated navigation and mid-water inertial-based integrated navigation.1.Based on the available navigation information of a HOV,the near-seabed navigation algorithm combining dead reckoning(DR)and ultra-short baseline(USBL)is studied.In order to improve the reliability of the DR/USBL integrated navigation system and reduce the design difficulty of information fusion,it is necessary to preprocess the USBL positioning data containing a large number of outliers and a small number of missing values,so that a common Kalman filter can fuse DR and USBL effectively.In order to obtain continuous and consistent positioning results,an online USBL data cleaning method based on online support vector machine regression(OSVR),whose feasibility and effectiveness are verified in aspects of accuracy and real-time,is proposed.Experiment results show that the DR/USBL integrated navigation system can effectively increase the data update rate and output smoother positioning results.2.In view of relatively high cost and difficulty in application of auxiliary navigation information sources(LBL and USBL)in midwater navigation,and the low accuracy or limited use in mid-water of DR,single-beacon added Strapdown Inertial Navigation System(SINS)method,from the perspective of reducing cost and difficulty of navigation system,was studied.In order to reduce the difficulty of observability analysis,a simple and intuitive analytical method of observability analysis is studied.The simulation results under static and horizontal circular motion are consistent with the analytical observability analysis results.The simulation test results show that the positioning accuracy of the single-beacon added SINS navigation system is related to the horizontal distance between the target and the beacon,and under the condition that the horizontal distance remains the same,changing the maneuvering mode of the carrier can improve the observability of navigation system.3.Aiming at the problem of insufficient available navigation information sources in midwater,Self-Aided SINS navigation algorithm for spiral dive/float HOV is proposed in order to further mine potential available navigation information sources.In order to improve the effect of digital filtering,a delay correction-high-pass filter(DC-HPF)was designed,and the simulation test has verified its superiority.In order to verify the feasibility and effectiveness of the Self-Aided SINS navigation algorithm,mathematical simulation and vehicle-mounted tests were carried out.The test results show that Self-Aided SINS only requires inertial measurement unit(IMU)as a sensor to obtain navigation and positioning results whose errors do not accumulate over time.4.Aiming at the application problem of doppler velocity log(DVL)working in watertrack mode when ocean current velocity is unknown,a noval SINS/DVL integrated system,based on Self-Aided SINS navigation algorithm,is proposed for midwater navigation by estimating ocean current velocity in advance.In order to ensure the realization of the algorithm,the navigation task is divided into two working modes: alignment mode and navigation mode.The ocean current velocity and installation error angle are estimated in the alignment mode,and the SINS/DVL/ocean current integrated system is realized in the navigation mode.The simulation test results show that the algorithm can effectively estimate the horizontal ocean current velocity and the heading installation error angle,and can obtain high-precision navigation results. |
