Research On Underwater In-motion Alignment For DVL-Aided SINS

For Autonomous Underwater Vehicle （AUV）, in order to guarantee itsindependence, From a cost point of view, the Strapdown Inertial Navigation System（SINS） aided by Doppler Velocity Log （DVL） is generally selected as the mainintegrated navigation system. For the application of SINS/DVL integrated navigationsystem of AUV nowadays, there are mainly two problems as following: one is theunderwater environment is changeable, and is vulnerable to interference such as thewaves, the other is DVL can’t get accurate velocity of vehicle under the condition oflarge water depth. All the above would influence the process of initial alignment ofinertial navigation system. Our research is on autonomous underwater vehicle, researchbased on the medium accuracy of strapdown inertial navigation system and integratednavigation system of Doppler log, we study the method of initial alignment of SINS/DVLintegrated navigation system on swaying and the moving base. Several works are carriedon as follows:（1）We study the principle of inertial navigation system, the solving algorithm andthe initial alignment method of SINS, and then we study the principle and the errorcharacteristic of DVL. At last we design the path generator of SINS and build the errormodel of SINS on Matlab/Simulink for the use of verification of algorithm.（2） Under the condition of swing base, we design the adaptive Kalman filter tocomplete initial alignment of inertial navigation system, to enhance anti-jamming,environmental adaptability of alignment method.（3） For moving base, DVL is likely to unlock of sea ground, so we adopt doubleprogram for initial alignment, the first one introduce the velocity of DVL as the reference,the second one we don’t introduce the velocity of DVL. And we use misalignment angledifference produced by the two programs to estimate DVL velocity error, and in the finalfine alignment, we use the estimation to compensate for the DVL velocity error, thesimulation results show that the proposed method can get more stable and higherprecision of initial alignment on moving base.