Gravity Modeling in GNSS-Aided Inertial Navigation System Safety Certification

Safety certification of GNSS-aided inertial navigation systems (INS) in civil aircraft requires thorough testing to ensure proper operation, even in worst-case conditions. One error that must be considered is that of gravity compensation on accelerometer measurements. Prior to the work described in this paper, no stochastic models existed with the Gaussian bounding of the tails required to ensure integrity performance. This paper describes a method to determine efficient stochastic models of the error of current high-order gravity models such as EGM2008. The stochastic and high-order models are combined to achieve a high-fidelity model suitable for use in testing systems designed for low-approach operations such as RNP-AR. This paper also describes a method to determine efficient stochastic models for low-order gravity models such as the WGS-84 ellipsoidal model. Such models may be used in testing systems designed for operations with less stringent lateral requirements.