• unmanned ground vehicle
• obstacle avoidance
• path planning

This thesis presents a novel algorithm for real-time path planning that can avoid both static and dynamic obstacles. The task is addressed as an optimal control problem, subject to a set of constraints. The algorithm is designed for an unmanned ground vehicle modeled with a non-linear dynamics and equipped with a LiDAR sensor that generates a sequence of point clouds. The algorithm implements a local planner, since the system is able to work in unknown enviroments. The information acquired from the LiDAR sensor is processed with the aim of deriving a 2D model of the obstacle. Then, a number of obstacle's points are evaluated by the planner as constraints. The novel contribution of this work relies in the obstacle term considered in the cost function to be minimized. The proposed algorithm is based on the knowledge of the system's dynamics, which is employed to predict the future state of the vehicle during path planning. The proposed approach has been tested both in simulated and real scenarios.