Gyrocompass Alignment with Base Motions: Results for a 1 nmi/h INS/GPS System

The accuracy of inertial navigation sensors (INS) in an integrated INS/GPS system is frequently specified by performance during gyrocompassing. Even though gyrocompassing is usually performed when the vehicle is on the ground, the base of an INS/GPS system is constantly subjected to random vibrations. The alignment problem for a 1 nmi/h class INS/GPS system with base motions is formulated as a 144-state Kalman filter with zero integrated velocity updates. Its base motions are modeled as second-order Markov stochastic processes. Results demonstrate that the alignment accuracy is extremely sensitive to base motions. Leveling errors converge very quickly, and their final values are generally not sensitive to initial estimates. On the other hand, azimuth alignment errors converge much more gradually. Error time histories are presented for a number of cases to show the interrelationship of important factors such as latitude, inertial sensor errors, measurement noises, and disturbance characteristics.