Simulation Environments for Real World testing

In this thesis the use of a simulation environment for the calibration of a path following robot controller is proposed. The controller modeled is a system used by Porsche Engineering for the execution of ADAS tests in the Nardò Technical Center facility in the south of Italy and has proven to be reliable under the correct tuning of its parameters. Since the tuning procedure has a high cost in terms of track usage and required human resources, the idea of a simulation aided tuning process is investigated. Firstly, the premises of vehicle dynamics will be introduced, in order to present the theoretical ground base from which to start. Then, the vehicle model structure and logic will be described, with the architecture of all the components utilized to represent its characteristics along with the longitudinal and lateral controller model structure and control strategies. After that, the validation processes of such models, and the obtained results, will be shown, and the proposed tuning strategy will be described. The final results will show several improvements in the modeled vehicle control.