Tóth Szilárd Hunor

Automated vehicle motion planning and control at handling limits with reinforcement learning

Budapesti Műszaki és Gazdaságtudományi Egyetem

Kandó Kálmán Doktori Iskola

Viharos Zsolt János

egyéni

PhD

One of the main drivers of research and development in the automotive industry is automated driving. In the short term, the manufacturers are focusing on the commercialization of advanced driver assistance systems. In the long run, the aim is to make available self-driving cars with SAE level 5 capabilities for a board range of society. A key element in an automated vehicle is its motion planning- and control system, which enables the vehicle''s movement in a well-predicted and accurate way in cooperation with an appropriate actuator and sensing system. This task can be challenging even for a professional driver if one of the vehicle tires reaches its saturation limit, see, e.g., special motorsport driving techniques like drifting, trail braking, pendulum tum, etc. Moreover, rapidly changing conditions, like spotted ice roads, aquaplaning, or beating up the tires, can also be hard to handle by a human. Classical control systems are usually based on models containing preliminary collected information about the vehicle and environment, moreover, they can be augmented with different types of observers. However, several aspects are hard to handle by modelling, just to mention the tire characteristics with uneven temperature over the profile, expected behaviour of vehicles (see, e.g., load changes, actuator dynamics, or battery state of charge variations), and uncertain or unforeseen environmental conditions. For such reasons, the use of learning algorithms in a standalone or combined way with classical control techniques is promising. The goal of this Ph.D. topic is to develop reinforcement learning algorithms, and control algorithm combinations that can precisely handle the vehicle motion dedicated at the handling limits and rapidly changing conditions and the could provide the maximal flexibility or vehicle control for safe behavior.

2023-09-01

2027-08-01

folyamatban lévő

Automated vehicle motion planning and control at handling limits with reinforcement learning