Implementation of Q Learning and Deep Q Network For Controlling a Self Balancing Robot Model

TL;DR

This paper explores implementing Q Learning and Deep Q Network algorithms to enable a self-balancing robot model to learn optimal actions for maintaining balance through reinforcement learning experiments.

Contribution

It presents the practical application and comparison of Q Learning and DQN on a simulated self-balancing robot, demonstrating their effectiveness in learning balance control.

Findings

Q Learning and DQN successfully learned balancing behaviors

Different parameters affect the learning efficiency

Plots illustrate the learning progress over time

Abstract

In this paper, the implementation of two Reinforcement learnings namely, Q Learning and Deep Q Network(DQN) on a Self Balancing Robot Gazebo model has been discussed. The goal of the experiments is to make the robot model learn the best actions for staying balanced in an environment. The more time it can stay within a specified limit , the more reward it accumulates and hence more balanced it is. Different experiments with different learning parameters on Q Learning and DQN are conducted and the plots of the experiments are shown.