Underwater Integrated Navigation Technical Research Based On SINS/DVL For Long-endurance Navigation

The ocean is an important strategic space for human development.In-depth exploration of the ocean has become one of the development themes of this century.Underwater vehicles are the main tools for humans to explore the ocean.Accurate underwater navigation positioning is the guarantee for the normal and stable operation of underwater vehicles,especially for underwater vehicles that work underwater on long-endurance.Since underwater vehicles operate underwater,many navigation methods cannot be applied to underwater vehicles such as optical navigation,radio navigation,satellite navigation,etc.The combination of the Strapdown Inertial Navigation System(SINS)and the Doppler Velocity Logger(DVL)is often used to provide navigation and positioning information for underwater vehicles.However,the SINS error will accumulate over time.The longer the time,the greater the error.Therefore,for long-endurance navigation systems,DVL can provide speed information to provide measurement signals for the navigation system,which can effectively reduce system errors.In this paper,the velocity information provided by DVL in long-endurance SINS/DVL combined system includes large disturbance noise and short-time failure of DVL is discussed to improve the accuracy of SINS/DVL navigation system based on long-endurance,providing reliable navigation data for underwater vehicles.Firstly,the principle analysis of the Strapdown Inertial Navigation System is carried out,and a complete mathematical model is established.Based on this,the error model is derived.The Doppler Velocity Logger is introduced to the principle of speed measurement,the cause of the error is analyzed,and the error model is derived.The two systematic error models established separately will lay the foundation for the integration of the two systems.Then,the Kalman filter equation and measurement equation of SINS/DVL integrated system are established,and the system is simulated numerically.It is verified that the position error and velocity error of SINS/DVL integrated navigation system can be effectively compensated by Kalman filter and feedback correction,so that the integrated navigation system can achieve certain positioning accuracy.The reason for the DVL misalignment is analyzed,and it is determined that the DVL velocity information contains large disturbance noise and the short-term failure of DVL is the main reason for its misalignment.Thirdly,an improved adaptive filtering algorithm is proposed for DVL output data with large disturbance noise.In order to solve the problem that the innovation covariance deviates from reality seriously when the disturbance noise of DVL becomes larger,the prior state mean square error matrix is adjusted adaptively by using the inaccurate innovation.The simulation verification of the situation is carried out.The improved adaptive filtering algorithm and the improved Sage-Husa filter are compared with the traditional Kalman filter.The simulation results show that the improved adaptive filtering can be well adapted to the DVL velocity disturbance noise.The situation can thus improve the accuracy and reliability of the long-endurance SINS/DVL integrated navigation system.Finally,in order to ensure the local navigation accuracy of the long-endurance SINS/DVL combination system when the DVL fails for a short time,this paper introduces Partial Least Squares Regression(PLSR)and Support Vector Regression(SVR)into the integrated navigation system,auxiliarying SINS/DVL navigation system.The PLSR and SVR are used to form a hybrid predictor.When the DVL is normal,the SINS and DVL output speed information can be trained to establish a mapping relationship between the SINS solve speed and the DVL output speed.When DVL fails,the predictor can replace DVL to output the measurement information needed by integrated navigation system in a short time,so as to ensure the local navigation accuracy of the system when DVL fails in a short time.The effectiveness of the proposed method is verified by simulation analysis.It can improve the accuracy and reliability of the system.