Autonomous driving framework for all-terrain vehicles using hyper fuzzy logic model

Autonomous driving in unstructured areas remains a challenging task for All-Terrain Vehicles (ATVs) due to the uncertainties coming from unpredictable environment conditions and sensor measurements. This paper proposes an Intelligent Driving System (IDS) based on a Type-2 Fuzzy Logic Model that integrates semantic reasoning and numerical sensor data to support adaptive decision-making response in situations where high levels of uncertainty are presented. The proposed IDS framework utilizes the Hyper Fuzzy Sets to cope with imprecision in spatial input information and output alternatives. The effectiveness of the proposed system is verified by simulating off-road scenarios over seven different types of terrains in several operational states. Experiments confirm that the proposed system can achieve an adaptive response for different terrain conditions while allowing for smooth and robust transitions between various output states. Such prominent performance makes the proposed system is a desirable option for autonomous off-road navigation in realistic and complex environments.