Eventos Anais de eventos
COBEM 2023
27th International Congress of Mechanical Engineering
Experimental Methods for Identifying Physical Parameters of Four-Wheeled Autonomous Vehicles
Submission Author:
João Bezerra , RN
Co-Authors:
João Bezerra, Felipe Wyatt Varga, Ely Carneiro de Paiva, Niederauer Mastelari, Rafael de Angelis Cordeiro, mauro ferreira koyama
Presenter: Niederauer Mastelari
doi://10.26678/ABCM.COBEM2023.COB2023-2060
Abstract
In the recent years, there has been significant research attention dedicated to robotic vehicles, both in Academy as well as in Industry. An important class of mobile robot is the Four-Wheel Autonomous Vehicle that has emerged as a disruptive technology with significant potential to revolutionize the transportation industry. The research efforts in this field, in the last decade, aim to develop strategies ranging from the development of Advanced Driver Assistance Systems (ADAS) up to the creation of fully autonomous navigation strategies. In this context, the mathematical models of the four-wheeled robots are an essential part of developing control techniques and autonomous navigation. Whether using complete mathematical models that accurately describe vehicle movement or simplified versions of these models that are used in the synthesis of controllers, we must provide the physical parameters of the platform in a more accurate way for these models. Without accurate parameters, the resulting controllers may be ineffective, leading to poor system performance and potentially even system instability, which can lead to unrealistic simulations that do not reflect the behavior of the actual system. This paper presents experimental methods to obtain some of the main physical parameters that determine the dynamic model of an autonomous vehicle, including its mass, its center of gravity, and the principal moments of inertia (in all three axes). The moments of inertia are measured using pendulum-type experiments, whose oscillation frequency is acquired through an inertial measurement unit (IMU) to determine the angular displacements. This paper also presents a cornering stiffness estimation method based on the vehicle lateral dynamics and an experimental dataset. The cornering stiffness of the wheels is related to the force models developed in the contact tire-ground, as well as the steering system dynamics. The experiment to estimate the cornering stiffness is based on the imposition of a chirp function to excite the lateral dynamics of the vehicle, where the IMU is used to yield the lateral velocity and the angular velocity signals, necessary for the estimation algorithm. The methods used in this work were applied to identify the vehicle parameters of the VERDE Project, whose platform is a four-wheeled 1:5-scale electrical vehicle that has an electronic differential in the rear for optimized torque/speed distribution. The achieved results demonstrate consistency with results reported in the related literature, and will be used to validate the VERDE dynamic simulator.
Keywords
Dynamical Models, Vehicle Dynamics, Cornering Stiffness, Experimental Methods, Parameters Determination
