Research On Integrated Navigation Algorithm Of SINS/DVL Based On Adaptive Filtering

Strapdown inertial navigation system relies on its own inertial sensing device to independently measure navigation information,and then navigates the computer to do inertial navigation calculation to determine the navigation parameters of the carrier.In recent years,SINS has the advantages of small size,high reliability,autonomy and good concealment,has become a research hotspot in the field of navigation,but also restricted by the principle of SINS defects,namely the internal system error will be accumulated with the increase of long work,so it can meet the high precision short-term navigation.In view of the problem of system error accumulating with time,scholars have adopted a variety of integrated navigation methods to solve the problem.At present,the mainstream is the combination of inertial navigation and GPS.But GPS has poor autonomy,the signal is susceptible to interference and the right to use such shortcomings so this decision by heteronomy,SINS and Doppler log combination method.In this paper,the advantages and disadvantages of GPS and DVL are analyzed,and the advantages of DVL are clarified,and the combination method of SINS/DVL has the advantages that SINS/GPS combination does not possess.Introduces the basic principle of DVL and SINS respectively,focuses on the four beam Janus configuration DVL measurement principle and its advantages and disadvantages,and then gives the performance,speed error model,DVL SINS,and according to the two kinds of model,system model of SINS/DVL integrated navigation.Secondly,according to the SINS and DVL between the lever arm and the space information inconsistency problem,analyzes the mechanism of lever arm effect produced by SINS and DVL will give way to a unified information space;for how the expansion into the system filter model of lever arm size error,do rigorous formula derivation,also clarifies how to apply the lever arm error estimates to compensate the error of navigation information,and through simulation experiments of this part were verified by simulation.Then,aiming at some defects,the classical Kalman filtering solution exists in the process such as: unknown noise statistics or system model is not precise enough case will lead to failure of the filter may choose to introduce the Sage-Husa adaptive filter to compensate for the lack of classical filter;adaptive filter estimation solution,because the system noise and the measurement noise matrix on-line estimation,there are large computing cost,the noise estimation algorithm in the presence of a negative criterion of instability and the resulting filter filter have the disadvantage of using this strategy,based on adaptive filtering to optimize the design,and to verify the correctness of the filtering algorithm optimized by simulation.Finally,we extract the DVL noise of the ship in mooring and dynamic state based on the existing sea trial data,and do the hardware in the loop simulation with the simulated data of the trajectory.The rationality of the algorithm is verified by multi group simulation experiments and comprehensive comparison analysis.