Towards the Design of Model-Based Path Tracking Vehicle Control Aware of Tyres’ Thermal and Wear State

Current Autonomous Vehicles (AVs) guidelines require increasingly stricter safety standards and aim towards developing sophisticated control logics to achieve multiple objectives in various scenarios. Indeed, major attention has been oriented towards meeting safety standards in high-velocity profile contexts where the reduction of accidents is closely related to the design of more accurate vehicle control logics. However, as the vehicle represents a intrinsically dynamic mechanical system, its operating (mass, tyre thermal and wear state) and external (weather) conditions continuously change and the performance of the controller can rapidly degrade, leading to increased safety-related risks and decreased comfort perception. This is mainly due to standard controllers not sharing their parameters with the vehicle since their calibration is usually performed in pre-defined conditions. To assess the impact of varying tyre and vehicle-related conditions, the path tracking scenario has been considered as it currently represents one of the most representative objectives when designing AVs control logics. For that purpose, a standard five-states lateral bicycle vehicle model with a Magic Formula (MF) model for tyre forces evaluation, has been employed in a Model Predictive Controller (MPC) framework with front steering as control signal. To the purpose of describing tyre wear state, a dedicated model considering the necessary working and boundary conditions, developed by the research group, has been considered. Various highway trajectories, at different speeds and road curvatures, have been simulated on a high-fidelity reference model of the real vehicle, employed as plant model. Simulations have been carried out by means of primary and secondary metrics: the former with the aim of evaluating path tracking performance through vehicle dynamics safety- and comfort-related indexes; while the latter employing tyre wear model to assess the degradation and overall environmental impact of the designed control logic.