INS Alignment Improvement Using Rest Heading and Zero-Velocity Updates

In this paper the feasibility of rapid alignment and calibration of a static strapdown inertial navigation system (INS) is evaluated. Resting conditions including zero-velocity update and a known initial heading direction as virtual external measurement data are integrated with INS data. By comparing the virtual external measurements with the estimates of those generated by the aligning INS, estimates of the velocity and heading errors can be obtained and these errors will be propagated in the INS as a result of alignment inaccuracies. An extended Kalman filter based on an augmented process model and a measurement model is designed to estimate alignment attitudes and biases of inertial sensors. Monte Carlo simulation results show that the integration of INS with rest conditions is very effective in rapid and fine leveling and azimuth alignment of INS, but this type of data fusion due to poor acceleration and angular rates of static condition has no chance of valuable calibration of all inertial sensor biases.