Real-Time Road Bank Estimation With Disturbance Observers for Vehicle Control Systems

A novel real-time road bank estimation method is proposed and experimentally verified by using a state-constrained Kalman-based disturbance observer on the combined lateral and roll dynamics of the vehicle. Road-body kinematics, which relates road angle rates and vehicle pitch/roll rates, is also implemented to improve the accuracy of vehicle angle estimates. The stability of the disturbance observer with known and stochastic initial conditions is investigated. The developed estimator uses data from inertial sensors available in driver-assistance and automated driving systems and does not require additional measurements. The main advantages of the proposed method are road angle estimation in maneuvers on pure and combined banked/graded roads and robustness against road disturbances. Several experimental tests include lane change on roads with different bank angles, with dry and slippery surface conditions, to thoroughly validate the algorithm and demonstrate its performance.

Automated summary

This paper proposes a novel real-time road bank estimation method, which is experimentally verified by using a state-constrained Kalman-based disturbance observer on the combined lateral and roll dynamics of the vehicle. The method improves the accuracy of vehicle angle estimates and has advantages in estimating road angles in maneuvers on pure and combined banked/graded roads with robustness against road disturbances.