Research On AUV Integrated Navigation System And Information Fusion Technology

As an important means for exploring ocean, underwater vehicle is widely applied incivilian and military. With the continuous improvement of requirements for navigationaccuracy, high-precision navigation is technical support for underwater vehicle fulfillingunderwater navigation tasks reliably and safely, Meanwhile, face to the complex underwaterwork environment, single navigation method has serious limitations in the application ofunderwater vehicle. Therefore, in view of the work process and actual applicationrequirements for underwater vehicle, this paper is researched on AUV integrated navigationsystem and information fusion technology, the major research work is as follows:1、In view of that underwater vehicle cannot quickly fulfill self-alignment, AUV movingbase initial alignment technology based on SINS/GPS is researched. First, based on Eulerplatform error angle, the nonlinear error model of strapdown inertial navigation system with alarge azimuth misalignment angle is established. Secondly, adaptive robust CKF algorithm isproposed to overcome that system will be affected by observation gross errors and inaccuratenoise statistical characteristics. Finally, According to simulation analysis and experimentalverification, adaptive robust CKF algorithm not only can improve the filter estimationaccuracy and convergence rate, but also can augment the stability and adaptive capacity, andwhich can meet the requirement of underwater vehicle moving base initial alignment.2、In view of the inconvenience of actual use and maintenance for traditional calibrationmethod, underwater vehicle gyro error on-line calibration technology based on SINS/CNS isresearched. First, gyro error on-line calibration model based on quaternion error is established.Secondly, in order to solve the problem that high-dimension system will make real-timereduced, according to the observable degree analysis of system state variable, and the mixedfilter algorithm combined by model prediction filter and recursive filter is proposed. Finally,according to the simulation analysis, the feasibility and superiority is verified.3、In view of the limitation that GPS and other navigation equipments communicating byradio cannot apply in water, AUV SINS/DVL integrated navigation technology is researched.First, SINS/DVL mathematical model is established, the affection of different correctionmethods on filter is analyzed, and the superiority of mixed correction method is verifiedaccording to the simulation. Secondly, in AUV actual navigation work, in order to solvesystem measurement noise may be unknown or time-varying and errors may exist in systemmodel, simplified Sage-Husa adaptive Kalman filter and estimation variance adaptiveweighting Kalman filter are proposed. Finally, according to the simulation analysis and experimental verification, the filter algorithm above can improve filter flexibility and restrictfilter divergence, and can on-line estimate the unknown or time-varying system measurementnoise and improve AUV SINS/DVL navigation accuracy.4、In view of the high-accuracy navigation requirement and working characteristics,remote AUV integrated navigation technology based on SINS/GPS water correction processis researched. First, SINS/GPS mathematical model is established. Secondly, in order toovercome the affection of surge, gusts and motivational variation, and to solve the problemthat GPS in high dynamic circumstance may appear measurement wild value, trackingadaptive Kalman filter algorithm with fault-tolerant is proposed. Finally, according to thesimulation analysis and experimental verification, the algorithm can improve navigationaccuracy in actual application.5、In order to fulfill the telling fusion of multi-sensor information and to further improveunderwater vehicle navigation accuracy, AUV information fusion technology based onSINS/DVL/MCP/TAN is researched. First, the corresponding integrated navigation systemmathematical model is established, based on PWCS and singular value decomposition,SINS/DVL integrated navigation system, SINS/DVL/MCP integrated navigation system, andSINS/DVL/MCP/TAN integrated navigation system are carried on observable analysis, andthen the best integrated method can be obtained to improve filter estimation capacity.Secondly, in AUV actual application, the measurement noise of navigation sensors may becolored noise, and that the determination of information distribution coefficient will affect theperformance of federal filter, in view of above problems, adaptive federal H_∞filteralgorithm based on developed Elman network is proposed. The simulation results show thatthe algorithm can meet AUV actual navigation requirement.