کنفرانس بین المللی کنترل ، ابزار دقیق و اتوماسیون

صفحه اصلی / هشتمین کنفرانس بین المللی کنترل ، ابزار دقیق و اتوماسیون

Reinforcement Learning based Sequential Controller for Mobile Robots with Obstacle Avoidance

نویسندگان :

Sara Mashhouri¹ Mohammadali Rahmati² Yasamin Borhani³ Esmaeil Najafi⁴

1- دانشگاه صنعتی خواجه نصیرالدین طوسی 2- دانشگاه صنعتی خواجه نصیرالدین طوسی 3- دانشگاه صنعتی خواجه نصیرالدین طوسی 4- Fontys University of Applied Sciences - دانشگاه صنعتی خواجه نصیرالدین طوسی

کلمات کلیدی :

Mobile robot, PID, PD, Reinforcement Learning, Object Detection, Obstacle Avoidance, ROS, YOLO

چکیده :

In mobile robots, the design of a sophisticated control system including obstacle avoidance plays an important role. This paper has two main parts. First, an object detection algorithm based on YOLO-v4 with custom classes and using a depth camera in real-time has been implemented in Robot Operating System (ROS) to resolve robot obstacle avoidance for a mobile robot. Second, the robot is controlled sequentially through a predefined path while the path is divided into three main parts, each with its own specific controller. The optimum parameters of every controller are calculated by applying Reinforcement Learning (RL) techniques. The provided experimental results show the accuracy of the designed obstacle avoidance system. Moreover, the simulation results validates the proper performance of the designed sequentially control system.