Skip to content

This simulation is designed to verify the effectiveness of our proposed method "Reinforcement-Learning-Based Control Framework for Leader-Following Cascade Formation of Multiple Perturbed Surface Vehicles" by comparing its performance with other existing methods.

Notifications You must be signed in to change notification settings

HoangAnhMasterDev/Reinforcement-Learning-Based-Control-Framework-for-Leader-Following-Cascade-Formation-of-SVs

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

88 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Introduction

This simulation is designed to verify the effectiveness of our proposed method " (RL) Reinforcement-Learning-Based Control Framework for Leader-Following Cascade Formation of Multiple Perturbed Surface Vehicles" by comparing its performance with other existing methods:

  1. (FXESO) Fixed-time extended state observer-based trajectory tracking and point stabilization control for marine surface vessels with uncertainties and disturbances
  2. (AISMC) An enhanced tracking control of marine surface vessels based on adaptive integral sliding mode control and disturbance observer

Method

This is a cascade control scheme with two loops:

  • The outer loop is the high-level trajectory generation, which generates the trajectory reference for the inner loop to achieve the target formation.
  • The inner loop is the low-level RL-based optimal controller, which minimizes the performance index.

Control strategy diagram

Result and Comparison

To begin with, we have the trajectory of all five SVs achieving the target formation. The picture presents snapshots of the simulation at different time instants.

Description

Subsequently, the following pictures depict the comparison of the control input in the x-axis, y-axis, and yaw angle of SV 4, from three different control schemes (AISMC, FXESO, and RL, respectively):

AISMC FXESO RL

Then, we have the tracking error performance in the x-axis, y-axis, and yaw angle, respectively of SV 4. It is observable that our control strategy is the only method that can guarantee the prescribed performance:

AISMC FXESO RL

Finally, we present the evolution of both Actor & Critic training weights of SV 4:

AISMC FXESO

It should be noted that we only show the results collected from SV 4 because the results of each SV are similar to those of other SVs.

Step-by-step guidance

  • FIRST, all the m files and simulink files are in MATLAB R2023a version. Make sure you have at least the R2023a version for the best performance.
  • Run AISMC simulation, remember to change the direction in the "save as txt file" section to your desired direction.
  • Run FXESO simulation, also remember to change the save-data direction
  • To run Actor-Critic simulation, run Parameters.m. Then, to have the data in txt file, run convertData2txt.m.
  • Now, run drawGraph.m to have graph.

Contact

Feel free to contact me via [email protected] if you need anything related to this repo.

About

This simulation is designed to verify the effectiveness of our proposed method "Reinforcement-Learning-Based Control Framework for Leader-Following Cascade Formation of Multiple Perturbed Surface Vehicles" by comparing its performance with other existing methods.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages