Robust aspects in dependable filter design for alignment of a submarine inertial navigation system

The strapdown inertial navigation system (SINS) is able to provide continuous estimates of a vehicle??s velocity, position and attitude. As a rule, the SIMS component known as a high accuracy strapdown inertial measurement unit (SIMU) is an exceptionally expensive system. Less expensive SIMUs comprised of low cost sensors suffer from degraded performance, but this can be compensated for, in part, by addition of a velocity data recorder (VDR) to accompany the SINS. In this configuration, the need frequently arises to align the SINS of a submarine to in order to avoid a long run-up of the inertial system before a start command is issued. This in-motion alignment (IMA) can be accomplished by integrating SINS data with some external aiding source, such as the VDR, by using some form of measurement matching method. Accordingly, this paper demonstrates a consistent IMA scheme for a low-cost SIMU using a robust Kalman/ $ H_{\infty } $ filter structure. An error model of the SINS is derived in which the state vector includes attitude, velocity, position and sensor errors. Velocity information from the VDR is used as a measurement to the proposed filter. All significant equations concerning navigation are presented in conjunction with argument. Results show the advantages of the approach and emphasize diverse aspects of the SINS.