Model-aided Strapdown Inertial Navigation System Integrated Method For AUV

SINS (Strapdown Inertial Navigation System) is used as the main navigation method for AUV (Autonomous Underwater Vehicle). In the absence of effective aiding, SINS will be divergent caused by error accumulation. DVL (Doppler Velocity Log) is often used to restrict its drift in this case. However, under certain circumstances, DVL detection range cannot reach the seabed because of the rough detection method, complex underwater terrain and other issues, and thus the feasibility of SINS/DVL integrated navigation mode may be reduced. The error of SINS increases rapidly when DVL fails to work and there is no accurate measurement.Also, the accuracy of Kalman filter is reduced by inaccurate system model and uncertain noise statistical properties. Thus, a feasible integrated navigation method is needed to restrict SINS velocity and position drift efficiently, and better robust filter is needed for the integrated navigation system state estimation. To solve the above problems, this paper presents a mathematical model aided SINS integrated navigation method for AUV based on fading memory Kalman filtering and H? filtering. The main contents of this paper are described as follows:Firstly, the differences between traditional integrated navigation method and model-aided integrated method are analyzed. The principles, mechanical layout, error analysis,and the error equations of SINS are introduced.Secondly, based on the motion model and the effect of sea current, mathematical model of AUV movement is establised. The position and velocity information of AUV is calculated with its three degrees freedom model, the combined external force and moment.Then, an improved fading memory Kalman filtering algorithm is proposed and applied to model-aided integrated navigation system. Uniform linear velocity motion and non-uniform velocity motion are simulated with accurate and inaccurate system model. Simulation results show that the accuracy of model-aided integrated navigation system is improved by the fading memory Kalman filtering method. And the Kalman filtering divergence with inaccurate model is inhibited.Finally, H? filtering algorithm with better robustness is applied to the model-aided SINS integrated navigation system. Uniform linear velocity motion and non-uniform velocity motion are simulated when the system model is accurate and inaccurate. Simulation results show that H? filtering method can improve the accuracy and robustness of model aided integrated navigation system. And the Kalman filtering divergence with inaccurate model is inhibited.The results in this paper show that model-aided SINS integrated navigation method for AUV based on improved fading memory Kalman filtering can inhibit SINS divergence and improve the navigation accuracy efficiently. Model-aided integrated navigation system based on H? filterin can effectively improve the system accuracy and robustness. The integrated navigation system proposed in this paper can work as a backup navigation system during DVL failure, and these two filtering methods can effectively restrict Kalman filtering divergence.