A Novel SINS/DVL Method Based on Smooth Variable Structure and Improved Variational Bayesian Adaptive Kalman Filter

Due to the complex and changeable underwater environment, the precision of SINS/DVL integrated navigation system is difficult to guarantee. Therefore, a novel SINS/DVL method based on smooth variable structure and improved variational bayesian adaptive Kalman filter is proposed in this paper. On the one hand, in order to reduce the influence of SINS/DVL navigation model errors, the sliding mode variable structure idea is introduced in this paper to calculate the optimal smooth boundary layer and constrain the influence of model errors. Then, a variational bayesian adaptive Kalman filter algorithm is proposed, which based on improved smooth variable structure. On the other hand, in order to reduce the influence of noise characteristics of unknown measurements on state estimation, an improved variational bayesian adaptive Kalman filter is proposed. The Sage-Husa adaptive noise estimation and strong tracking principle are introduced to improved VBAKF method. Finally, the simulation and river test are designed in this paper, while some other methods are compared.