Center of Gravity Estimation Using Multiple Accelerometers

An Estimation-Before-Modeling approach is utilized to estimate smoothed time histories of the center of gravity (c.g.) position from flight data for an ultralight helicopter. The helicopter is equipped with a flight test instrumentation system, consisting of multiple Inertial Measurement Units (IMUs) distributed at different positions across the vehicle. An Extended Rauch-Tung-Striebel (ERTS) smoother was utilized to reconstruct the c.g. position, together with smoothed time histories for external forces and moments acting on the bare airframe. This was achieved by incorporating all inertial measurements of the different IMUs into the estiamtion process. The solution to the stochastic flight path reconstruction problem obtained with the ERTS smoother is compared to a maximum likelihood-based approach with full discretization, considering the process noise terms in the overall optimization vector. The reconstructed c.g. time histories using the approaches mentioned above are verified using a simple least squares-based approach which relies only on kinematic relation between different IMU measurements.