Application of fixed point smoothing to the calibration, alignment and navigation data of inertial navigation systems

The authors present an application of the fixed-point smoothing technique to calculate the relative azimuth of inertial measurement units (IMU), by postprocessing of calibration and alignment data. It is shown that the smoother performs best when the filter and smoother error covariances for gyro fixed-drift-rate parameters are increased. A test scenario was created with the calibration and alignment data from an IMU simulator. The data were postprocessed and the smoother estimated all but 0.1 arc seconds of the filter azimuth error. This is well within the expected accuracy for relative azimuth computation