Identification of Vehicle Mass Via Experimental Data Analysis Under Real Driving Conditions for On-Board Advanced Mobility Technologies Applications

The growing emphasis on road safety and environmental sustainability has spurred the development of technologies to enhance vehicle efficiency. Accurate vehicle mass knowledge is crucial for all vehicles, to optimize advanced driver assistance systems (ADAS) and CCAM (Connected, Cooperative, and Automated Mobility) systems, as well as to improve both safety and energy consumption. Moreover, the continuous need to report precisely on the greenhouse emissions for good transports is becoming a key point to certificate the impact of transportation systems on the environment. Mass influences longitudinal dynamics, affecting parameters such as rolling resistance and inertia, which in turn are critical to adaptive control strategies. Moreover, the knowledge of vehicle mass represents a key challenge and a fundamental aspect for fleet managers of heavy-duty vehicles. Typically, this information is not readily available unless obtained through high-cost weighing systems or estimated approximately through calculations that consider the average density of the transported load. This study presents an advanced methodology for vehicle mass estimation based on a longitudinal dynamic model, applicable to both light-duty and heavy-duty vehicles. The method was initially validated using experimental data from vehicles with known mass via the OBD system, then it was extended to heavy-duty vehicles through data extracted from the CAN bus following the J1939 standard. The reliability of the estimates was assessed by comparing them with benchmark values under various operating conditions for light-duty vehicles, whereas for heavy-duty vehicles, it was analysed using a statistical approach to account for the lack of precise vehicle payload knowledge. Additionally, a sensitivity analysis assessed the influence of uncertain parameters, such as the aerodynamic drag coefficient and the powertrain efficiency. The results show a satisfactory agreement of the estimated masses and confirmed the straightforward use of the methodology for mass estimation to improve both dynamic performance and overall vehicle efficiency, contributing to road safety and reducing pollutant emissions. This study provides a significant contribution to the field of sustainable mobility by offering a robust and scalable method for the automotive industry. It is a low-cost approach that requires only a limited set of experimental data and can also be integrated into more advanced models. The proposed methodology can support the development of autonomous driving systems, contributing to safer and more adaptive mobility solutions.