Adaptive Factor Graph-Based Tightly Coupled GNSS/IMU Fusion for Robust Positionin

2025-11-28

AI Summary

To address insufficient localization robustness under GNSS-challenged conditions (e.g., urban canyons), this paper proposes a tightly coupled GNSS/IMU factor graph optimization framework. Methodologically, it jointly fuses pseudorange measurements and IMU pre-integration factors, and introduces a novel adaptive Barron loss function with a unified, tunable parameter to dynamically suppress non-Gaussian noise and outliers within the factor graph—automatically down-weighting unreliable GNSS measurements. The framework is efficiently solved using an extended GTSAM implementation. Experiments on the UrbanNav dataset demonstrate a 41% reduction in positioning error compared to standard factor graph optimization, and significantly outperform conventional EKFs—especially in severely GNSS-obstructed regions. The core contribution lies in embedding an adaptive robust loss function into a tightly coupled GNSS/IMU factor graph, achieving a favorable trade-off among accuracy, robustness, and computational efficiency.

Technology Category

• Intelligent Robots: State Estimation
• Search and Optimization: Learning to Search
• Computer Vision: Learning & Optimization for CV

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Reliable positioning in GNSS-challenged environments remains a critical challenge for navigation systems. Tightly coupled GNSS/IMU fusion improves robustness but remains vulnerable to non-Gaussian noise and outliers. We present a robust and adaptive factor graph-based fusion framework that directly integrates GNSS pseudorange measurements with IMU preintegration factors and incorporates the Barron loss, a general robust loss function that unifies several m-estimators through a single tunable parameter. By adaptively down weighting unreliable GNSS measurements, our approach improves resilience positioning. The method is implemented in an extended GTSAM framework and evaluated on the UrbanNav dataset. The proposed solution reduces positioning errors by up to 41% relative to standard FGO, and achieves even larger improvements over extended Kalman filter (EKF) baselines in urban canyon environments. These results highlight the benefits of Barron loss in enhancing the resilience of GNSS/IMU-based navigation in urban and signal-compromised environments.

Problem

• Enhances GNSS/IMU fusion robustness against non-Gaussian noise and outliers.
• Adaptively downweights unreliable GNSS measurements for resilient positioning.
• Reduces positioning errors in urban and GNSS-challenged environments.

Innovation

• Adaptive factor graph integrates GNSS pseudorange with IMU preintegration
• Barron loss function downweights unreliable GNSS measurements robustly
• Extended GTSAM implementation reduces positioning errors in urban canyons