Autonómna jazda v simulátore

Abstract

This diploma thesis focuses on the development of autonomous vehicle control in the CARLA simulator. The theoretical part describes autonomous vehicles, their classification according to SAE, their development, and the sensors used. Furthermore, simulators used in the development of autonomous vehicles are discussed, with a detailed description of the CARLA simulator, including its architecture and applications. The thesis also includes an overview of machine learning methods used in the practical part. The practical part presents the design and implementation of a solution involving data collection, neural network design and training, and addressing challenges during learning. Finally, the results of testing are presented. The conclusion summarizes the achieved results and suggests possible future improvements.