Master's Thesis: Optimal Gait Control of Soft Quadruped Robots by Model-Based Reinforcement Learning

Overview

This repository contains the source code, data, and documentation for my master's thesis titled "Title of Your Thesis." This repository contains the code and documentation for the master thesis titled "Optimal Gait Control for Soft Quadruped Robots Using Model-Based Reinforcement Learning". The thesis explores the use of model-based reinforcement learning techniques to optimize gait control for soft quadruped robots, aiming to enhance their adaptability and efficiency across various terrains and environments.

Table of Contents

• Introduction
• Dependencies
• Usage
• Results
• License
• Contact

Introduction

The master's thesis investigates [provide a brief overview of the research problem, objectives, and methodology]. The study aims to [briefly describe the research goals and expected outcomes].

Dependencies

The project requires the following dependencies:

• Python 3.7 or higher
• pygame
• adafruit_bno055
• MATLAB 2021b
• Simulink with SimScape

Usage

To replicate the experiments conducted in the thesis, follow these steps:

1. Clone the repository: git clone https://github.com/n7729697/KTH-MasterThesis.git
2. Navigate to the RQ1/ directory: cd code/RQ1/
3. Run 'collect.m' to collect the samples
4. Run 'model_train_rq1.m' to train the DNN model
5. Navigate to the code/ directory: cd code/
6. Run 'MBRLmain.mlx' to train the agent

Results

The results of the experiments are presented in the img/ directory. The findings are summarized and discussed in detail in the thesis document available in the sections/ directory.

License

This project is licensed under the MIT License.

Contact

For any inquiries or feedback regarding the thesis or repository, please contact NIU Xuezhi at [email protected].