Performance analysis of a federated ultra-tight global positioning system/inertial navigation system integration algorithm in high dynamic environments

fei xie,jianye liu,rongbing li,bin jiang,li qiao

DOI: https://doi.org/10.1177/0954410014525359

2015-01-01

Abstract:The Doppler frequency changes rapidly due to high dynamics of vehicle, which leads to the loose lock and even the abnormal performance of global positioning system (GPS) receiver. To solve this problem, a federated ultra-tight integration algorithm based on pre-filters is proposed to optimal estimate both receiver tracking control commands and inertial navigation system (INS) navigation solutions. Firstly, the INS error model and GPS receiver tracking loop structure are built to present the fundamental architecture of the proposed ultra-tightly coupled system. Meanwhile, in order to reduce the load of the integrated filter, the pre-filters are incorporated to the ultra-tightly coupled system, and the state variables are fed into the integrated Kalman filter. Secondly, the intrinsic relevance between the phase and frequency biases of replica signals and INS states is analyzed to accomplish the deep fusion of INS and tracking loop. Finally, semi-physical simulations are performed by using a GPS signal simulator to generate signals of two high dynamic trajectories. The experimental results indicate that the proposed ultra-tight integration algorithm can achieve a good performance on reliable positioning and robust tracking in high dynamic environments, compared with the conventional approaches such as tightly coupled integration strategy and third-order phase-locked loops.