Quasi-Static Approach for Mass Estimation of Electric Propelled Vehicles

This paper presents an approach for estimating the mass of a vehicle by observing the oscillation behavior of the acceleration signal after a pulse excitation of the system. The short and controlled duration of the excitation pulse used by the vehicle's electric drive merely causes the vehicle to oscillate and does not result in a deeper rolling motion of the wheels, which means that the vehicle remains in a quasi-static range. The dependence of the oscillation on the mass results in a characteristic oscillation profile. A heuristic relation between vehicle mass and damped oscillation frequency for different loading conditions is generated, which can be used to estimate the vehicle mass. It is shown that the oscillation behavior is dependent on the load position and tire pressure of the vehicle. A simple and an extended physical model are proposed to reconstruct the experiments that were conducted beforehand. Subsequently, these models can be used to ensure an efficient process in generating the proposed heuristic relation. To ensure that the model is as realistic as possible, the excitation pulse of the drive train is simulated with the aid of a real-time Hardware in the Loop (HIL) emulator from Typhoon HIL. The electrical model contains a physical model of the inverter and electric machine and is combined with a SIMULINK model of the trailer. The parameter estimation of the unknown spring-damper constants of the real world system is solved by optimizing the model parameters with respect to the real oscillation signals that were measured on an electrified trailer for use in micromobility solutions.