Vehicle System State Estimation Based on Adaptive Unscented Kalman Filtering Combing With Road Classification

• Main Authors: Zhenfeng Wang, Yechen Qin, Liang Gu, Mingming Dong
• Format: Article
• Language: English
• Published: IEEE 2017-01-01
• Series: IEEE Access
• Subjects: State estimation; AUKF; process noise variance; measurement noise covariance; vehicle system
• Online Access: https://ieeexplore.ieee.org/document/8101477/

This paper presents a new method to address issues associated with vehicle system state estimation using an unscented Kalman filter (UKF) with considering full-car system and nonlinear tire force under various international standards organization (ISO) road conditions. Due to the fact that practical road information is complex and noise covariance cannot be treated as a constant, the influence of varying vehicle system process noise variance and measurement noise covariance on the estimation accuracy of the UKF is first discussed. To precisely estimate road information, a novel road classification method using measured signals (vertical acceleration of sprung mass and unsprung mass) of vehicle system is proposed. According to road excitation levels, different road process variances are defined to tune the vehicle system's variance for application of UKF. Then, road classification and UKF are combined to form an adaptive UKF (AUKF) that takes into account the relationship of different road process noise variances and measurement noise covariances under varying road conditions. Simulation results reveal that the proposed AUKF algorithm has higher accuracy for state estimation of a vehicle system under various ISO road excitation condition.