A Tightly Coupled Positioning Solution for Land Vehicles in Urban Canyons

Abstract

The integration between Global Navigation Satellite System (GNSS) and on-board sensors is widely used for vehicle positioning. However, as the main information source in the integration, the positioning performance of single- or multiconstellation GNSSs is severely degraded in urban canyons due to the effects of Non-Line-Of-Sight (NLOS) and multipath propagations. How to mitigate such effects is vital to achieve accurate positioning performance in urban canyons. This paper proposes a tightly coupled positioning solution for land vehicles, fusing dual-constellation GNSSs with other low-cost complementary sensors. First, the nonlinear filter model is established based on a cost-effective reduced inertial sensor system with 3D navigation solution. Then, an adaptive fuzzy unscented Kalman filter (AF-UKF) algorithm is developed to achieve the global fusion. In the implementation of AF-UKF, the fuzzy calibration logic (FCL) is designed and introduced to adaptively adjust the dependence on each received satellite measurement to effectively mitigate the NLOS and multipath interferences in urban areas. Finally, the proposed solution is evaluated through experiments. The results validate the feasibility and effectiveness of the proposed solution.