Eventos Anais de eventos
DINAME 2017
XVII International Symposium on Dynamic Problems of Mechanics
Analysis of control strategies for autonomous motorcycles - stabilization and trajectories tracking
Submission Author:
Marília Assad , RJ
Co-Authors:
Marília Assad, Mauro Speranza Neto, Marco Antonio Meggiolaro
Presenter: Marília Assad
doi://10.26678/ABCM.DINAME2017.DIN17-0041
Abstract
Autonomous vehicles - defined as vehicles with carrying capacity of persons or property without the use of a human driver - are an interesting and recent problem, with increasing studies in the last twenty years. Besides being able to position itself globally and to predict trajectories, these systems have to deal with obstacles and traffic signals - a common situation for a human driver, but it requires a combination of sensors and control techniques for these autonomous systems. Regarding this type of vehicles, a less explored option is the motorcycle. Apart from the difficulties inherent in making a vehicle move independently, autonomous motorcycles have to be able to remain stable at any speed and trajectory. This work’s main object of study is a small scale electric motorcycle, originally radio controlled. This paper presents a representative linear model of the motorcycle’s dynamics. The multibody approach is the most complete model of this type of vehicle’s behavior, where the motorcycle is interpreted as a combination of rigid or flexible bodies - including the wheels, bumpers, chassis, handlebar, fork and driver - which separately have a characteristic behavior and together influence the dynamics of each other. To decrease the complexity of this type of system, one alternative is to obtain a more simplified multibody model, in which only the main components of the vehicle are considered, such as the driver, wheels and front fork. This approach results in lower order models, which are easier to simulate and to apply classical or modern control strategies. The two-wheeled vehicle is considered as an inverted pendulum with a mobile base and other simplifications are proposed, as constant displacement speed or small steering and yaw angles. Since the vehicle chosen as object of this study is naturally unstable, fact proven empirically and mathematically, it is clear that, to ensure a follow-up course without overturning, it is necessary to apply an adjusted control signal according to the motorcycle's behavior. The human driver does this unconsciously, setting the course through the handlebars and/or tilting the body to avoid falling. Once the autonomous system studied will not have the presence of a mechanical counterbalance, there remains only the first choice as a control strategy. Thus, in this work, the dynamic characteristics of the zero track vehicles are analyzed and verified the validity of different stability and path tracking control strategies of a motorcycle using as input only the steering of the handlebar, and scale model real data.
Keywords
Autonomous vehicles, Motorcycle dynamics, Stability control, Trajectory control
