Automated Vehicle Attitude and Lateral Velocity Estimation Using a 6-D IMU Aided by Vehicle Dynamics.

Xin Xia,Lu Xiong,Wei Liu,Zhuoping Yu

DOI: https://doi.org/10.1109/ivs.2018.8500503

2018-01-01

Abstract:In this paper, an estimation method for attitude and lateral velocity for automated vehicle has been proposed using a six degree of freedom inertial measurement unit(IMU) aided by vehicle dynamics. This estimation method makes full use of the advantage of the IMU and vehicle dynamics and could run autonomously without aid from other outside information such as GNSS or camera. Based on Kalman filter theory, three observers have been developed: a kinematic model based attitude observer for pitch angle and roll angle using IMU, a kinematic model based lateral velocity observer for lateral velocity using IMU and a dynamic model based observer using vehicle dynamics. In small excitation condition, the estimated lateral velocity from dynamic model based observer is more reliable and it is forwarded to the two kinematic model based observers to prevent the accumulated estimation error. In larger excitation condition, the two kinematic model based observers run in open-loop mode. Slalom maneuver and double lane change(DLC) maneuvers have been conducted to validate the estimation method. The experiment results have proved the effectiveness of this estimation method.