Deterministic Particle Filter Based Transfer Alignment Method for Large Misalignment Angle

Based on multivariate distribution F-deviation representative point of number theory, an improved Deterministic Particle Filter (DPF) algorithm is presented in this paper, in which unequal weight is used for particle generation, important particle sampling and particle re-sampling. To meet the actual needs, a nonlinear transfer alignment model for Airborne Inertial Navigation System with large misalignment angle is constructed, which is based on multiplicative quaternion theory. By utilizing the DPF algorithm, a simulation for nonlinear transfer alignment model is performed. The accuracy and effectiveness of multiplicative quaternion theory based nonlinear transfer alignment model as well as high filtering precision of DPF are verified by simulation results. The proposed approach can be potentially used for the transfer alignment system design of Airborne Inertial Navigation System.