World models for autonomous driving

Advancements in autonomous driving have been achieved due to the increasing use of artificial intelligence. The existing autonomous driving technology is insufficient to handle intricate traffic conditions. In autonomous driving, the world model is a crucial and innovative technology. Utilizing sensors and pre-existing knowledge, a world model can enhance the autonomous driving system's comprehension of the surroundings, offer essential data for future judgments, and enhance the system's resilience. The paper employs literature analysis and review methods to investigate the research and implementation of world models in autonomous driving, including environment perception and modeling, path planning, decision-making, and safety. This study examines the use of artificial intelligence technology in autonomous driving and analyzes the research and application of the world model in this field. It offers new insights for addressing challenging scenarios in autonomous driving and enhancing the safety of the system.