Paper Title
LEARNING TO NAVIGATE: MOBILE ROBOT NAVIGATION WITH DDPG REINFORCEMENT LEARNING ALGORITHM AND CONTINUOUS ACTION SPACE

Abstract
Abstract - The use of reinforcement learning (RL) algorithms for mobile robot navigation has shown promising results. In this study, we focus on using the Deep Deterministic Policy Gradient (DDPG) algorithm for mobile robot navigation, with wheel speed as the action space. The DDPG algorithm is a model-free RL algorithm that can learn complex policies for continuous action spaces. The mobile robot is equipped with proximity sensors for environment perception. The DDPG algorithm is used to optimize the robot's policy and wheel speed values to navigate towards a specific goal while avoiding obstacles. The performance of the trained robot is evaluated in a simulated environment. The results show that the DDPG algorithm can effectively train a mobile robot to navigate towards a goal while avoiding obstacles using wheel speed as the action space. Keywords - Reinforcement Learning, DDPG, Coppelia Sim, Open AI Gym