A system identification approach to center of mass estimation in trucks

Abstract

The position of the center of mass of a vehicle is an important parameter as it significantly affects vehicle loads distribution and vehicle dynamics. This article proposes an approach to estimate the longitudinal position of the center of mass and the inertial parameters from offline batches. The proposed algorithm is based on the dynamical modelling of the vehicle and on the measurements of cheap gyroscopes and accelerometers. The estimation procedure uses the Box-Jenkins method for system identification to estimate the parameters given the noisy measurements. Prior knowledge of the vehicle is used to filter the output measurement and simplify the estimation problem. The approach is tested on simulations when one or more gyroscopes and accelerometers are used.Index Terms-System identification, vehicle dynamics, position of the center of mass, measurement noise, Box-Jenkins model.