A Stochastic Model-Based Fusion Algorithm for Enhanced Localization of Land Vehicles

• Published in: IEEE transactions on instrumentation and measurement Vol. 71; pp. 1 - 10
• Main Authors: Guo, Ge, Liu, Jiageng
• Format: Journal Article
• Language: English
• Published: New York IEEE 2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Adaptation models; Algorithms; Dead reckoning; Driving conditions; Estimation; extended Kalman filter (EKF); Kalman filters; Kinematics; kinematics model; Localization; Localization method; Location awareness; Mathematical models; Robustness (mathematics); sensor fusion; Stochastic models; Stochastic processes; Uncertainty; vehicle localization; Vehicles
• ISSN: 0018-9456; 1557-9662
• DOI: 10.1109/TIM.2021.3137566

This article investigates a position estimation problem for land vehicles using sensors fusion and dead-reckoning (DR) to mitigate the influence of model inaccuracy and uncertain noise covariance. The kinematics of the vehicle is roughly modeled, considering the roll angle and slip angle. To achieve accurate position estimation, a novel stochastic model-based fusion algorithm is proposed by embedding absolute value modulated random noises into the model. For uncertainties that are Gaussian, a quantitative description of the deviation due to uncertainties is given. Improved state and measurement equations are derived to enhance the accuracy of positioning. The algorithm recursively provides robust estimations in a stochastic manner. The effectiveness and superiority of the proposed vehicle localization method with inadequate process knowledge is demonstrated by numerical simulations and real-world experiments. Experimental results also demonstrate that our method is more accurate and reliable than the state-of-the-art methods for vehicle localization under various driving conditions.