A Hybrid A-Star Based Global Vehicle Planner for Uneven Terrains

This thesis was developed during an internship at Hipert Lab at University of Modena, laboratory specializing in software development and hardware implementation for autonomous driving across various types of vehicles. The work focuses on the study, development, testing, and simulation of a Global Planner based on the Hybrid A* algorithm, specifically designed for off-road and uneven environments. The main motivation of this project stems from the need for autonomous vehicles to safely and efficiently navigate complex terrains, where longitudinal and lateral inclinations significantly affect vehicle stability and trajectory feasibility. The goal is to design a Global Planner capable of planning a trajectory from point A to point B, taking into account the terrain slope in the evaluation of the path. The limited treatment of this topic in the existing literature posed significant design challenges. To address these, the project extended the standard Hybrid A* algorithm by incorporating methods for evaluating terrain slope and vehicle inclination. Different simulations were carried out to test and validate the planner under various environmental and operational conditions. The results demonstrate that the implemented Global Planner can generate feasible and safe trajectories while respecting a minimum overall pitch and roll values and the inclinations vehicle thresholds.