Assessment of a simple methodology for the identification of inertial parameters of semi-trailers

Road transport is the most common means of transporting goods in Europe. Heavy goods vehicles are largely employed for this sake and fully automated vehicles, even for heavy goods, are likely to play a role in the next few years. Within this context, models, and algorithms controlling the automated drive need to identify the mass and other inertia parameters related to the actual vehicle condition, which may vary during a mission (due to the transported goods arrangement). In this paper, a simple methodology based on an identification algorithm for the inertial parameters of a semi-trailer is presented, as well as an assessment of its result quality. Two analytical models of a semi-trailer truck were developed: a longitudinal model, allowing to identify the mass and the center of gravity (COG) position of the trailer, and a ride model, allowing to identify the roll and pitch moments of inertia of the trailer. The input data for the identification algorithm were obtained performing dynamic maneuvers on a multibody model; the model was validated in previous activities and provides, in this study, ground-truth (reference) signals. Different sets of inputs were considered, referring to the availability of different sensors on the actual vehicle. Based on such inputs, mass and inertia parameters were computed by the identification algorithm and were compared with the actual mass and inertia values. The error on the identification of the mass and the longitudinal position of the COG was always below 2% and the error on the identification of the vertical position of the COG was in the range 2%–10%, depending on the sensor configuration adopted. The error on the roll and pitch moments of inertia was about 25% using the baseline sensor configuration but reduces to 5% if a wider sensor set is used. • Identification of trailer inertial parameters including pitch and roll moments • Methodology based on simple vehicle dynamics models and least square method • Analysis of identification error of inertial parameters for different sensor setups