Optimized GNC Techniques for Service Robotics

In this thesis, we propose the development of an optimal control system for autonomous robots. Our design aims to efficiently guide the robot, determining the best possible route to its destination. We leverage the state-of-the-art Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm to direct the robot. By utilizing a precise navigation system, we can ascertain the robot's position in real-time and manage its movements by adjusting its components. This algorithm, which integrates principles from deep learning and reinforcement learning, offers superior optimization capabilities for robot navigation and control. Notably, our approach facilitates navigation optimization without relying on a pre-existing map and ensures collision avoidance throughout the journey.