Advances in MEMS IMU Cluster Technology for Small Satellite Applications

In recent years, there has been increased interest in Micro-Electro Mechanical Systems (MEMS) Inertial Measurement Units (IMUs) due to their relatively small volumetric footprint and low-cost. Although this advantage far outweighs the volumetric footprint and cost of traditional high-performance IMUs, MEMS technology has yet to match the performance of such devices. In spite of this, it has been shown in theory that a cluster of MEMS IMUs may signicantly improve the performance over a single MEMS IMU. To further develop this theory, two prototype boards have been designed and constructed that include 16 MEMS accelerometers and gyroscopes af- xed to a single Printed Circuit Board (PCB). To prove this technology, hardware and software has been developed for calibration and fault detection, which represents the majority of the body of this thesis. An apparatus has been designed to easily acquire three-axis measurements from the cluster prototype on a single-axis rate table. These measurements may then be placed into a Maximum Likelihood Estimation (MLE) algorithm in order to acquire the necessary error coecients incorporated in IMU measurements. Once these error coecients are accurately determined, future measurements may be calibrated. Finally, a fault detection, isolation, and recovery (FDIR) architecture was developed and simulated to determine faulty measurements in real-time, so that bad measurements may not be placed into downstream navigation lters. The hardware, software, and testing developed and performed in this thesis will be used in the verication process of an IMU cluster to help prove its worthiness in modern day small satellite applications