Fast IMU Autonomus Alignment, Gyroscopic and Accelerometer Measurement Improvement by Input Reconstruction

Navigation performance of Inertial Measurement Units (IMUs) is governed by the accuracies of initial alignment, gyroscopic drift and in the presence of fast maneuvers by the settling time of gyroscopes and accelerometers. The proposed SBIR work is a re-visit of the autonomous calibration problem using Higher Order Sliding Mode (HOSM) control solutions. The HOSM approach allows the separation of complex Multiple-Input, Multiple-Output control problem into a set of simpler Single-Input, Single-Output control problems. This approach achieves faster convergence and settling times, and is robust to system and gyroscope noise. HOSM observers will be developed, using internal dynamical models of the sensor electro-mechanical dynamics, to reconstruct (estimate) the input signal from the IMU measured output. Reconstructed signal settling time is solely governed by observer settling time and thus, is much smaller than sensor electro-mechanical dynamical response. Moreover, the reconstructed signal can be made practically devoid of measurement noise without adverse effects on the response time. This approach is similar to digital thermometers that produce temperature measurement estimates much before the thermometer thermal equilibrium is reached. The proposed work will benefit IMU accelerometer and gyroscope measurement performance and reduce autonomous alignment times, but is applicable to a wide class of measurement systems. Approved for Public Release | 16-MDA-8917 (15 November 16)