Moving Horizon Estimation for Real-time Dynamic Precise Point Positioning assisted by Orientation acquired from three-antenna POS system

Precise Point Positioning (PPP) typically experiences long convergence times, and Extended Kalman Filters (EKF) offers only limited robustness in challenging environments. Multi-antenna systems enable rigid body localization and vehicle orientation, making them a preferred choice for enhancing positioning accuracy without integrating additional sensors. In this work, we introduce orientation information to constrain velocity estimation, thereby assisting the PPP process. By fusing orientation data and exploiting the robustness of dynamic Moving Horizon Estimation (MHE), we propose an Orientation-Assisted Moving Horizon Estimation (OAMHE) algorithm to improve GNSS positioning performance, particularly in rough terrains with limited GNSS satellites. Meanwhile, we derive the constrained Cramér-Rao lower bound (CRLB) and calculation complexity. Finally, the field test results show that the OAMHE algorithm achieves faster convergence at the start of positioning and maintains stable performance without large errors even when GNSS satellite availability is reduced.