Vehicle Cooperative Positioning With Tightly Coupled GNSS/INS/UWB Integration Based on Improved Multiple Fading Factors and Adaptive Cost Function

Abstract

Cooperative positioning technology based on multi-vehicle information fusion is essential for advanced applications in intelligent transportation systems (ITS). The integration of global navigation satellite systems (GNSS), inertial navigation system (INS), and ultra-wideband (UWB) technology holds significant promise for enhancing the continuity and reliability of vehicle cooperative positioning. In tightly coupled GNSS/INS/UWB integration, the tolerance against measurement outliers and state model perturbations is pivotal for fulfilling the specific requirements of critical ITS applications. To optimize the comprehensive performance of vehicle cooperative positioning under uncertain sensor observation environments, this paper proposes a robust multiple fading factors unscented Kalman filtering (RMFUKF) algorithm based on adaptive cost function. The proposed solution incorporates Huber M-estimation with an adaptive tuning strategy to perform measurement-specific outliers processing. Furthermore, the improved multiple fading factors based on an exponential weighting method are implemented to mitigate the effects of dynamic model mismatches. Experimental results from vehicular field experiments demonstrate that the proposed RMFUKF scheme significantly improves the robustness and adaptive performance of vehicle cooperative positioning under unpredictable, real-world operating conditions.