Development of Multi-Agent Control for Surface and Underwater Vehicles

The effective deployment of autonomous amphibious vehicles for surface and underwater operations re- quires advancements in hardware, control systems, and decision-making algorithms to operate efficiently across dynamic environments. Current systems often lack seamless state estimation, multi-agent coor- dination, and adaptive control strategies for real-time obstacle avoidance in diverse operational settings. To achieve robust and adaptable performance, there is a need for a unified framework that integrates priority-based task assignment, state-aware control switching, and flexible leader-follower dynamics for collaborative operations among surface and underwater agents. The effective deployment of autonomous amphibious vehicles for surface and underwater operations requires advancements in hardware, control systems, and decision-making algorithms to operate efficiently across dynamic environments. Current systems often lack seamless state estimation, multi-agent coordination, and adaptive control strategies for real-time obstacle avoidance in diverse operational settings. To achieve robust and adaptable perfor- mance, there is a need for a unified framework that integrates priority-based task assignment, state-aware control switching, and flexible leader-follower dynamics for collaborative operations among surface and underwater agents.

Objectives

• Modeling of AUV, ASV:

  • Creating accurate models to represent both the autonomous underwater vehicle (AUV) and the autonomous surface vehicle (ASV) dynamics is crucial for simulating real-world behavior, enabling effective depth and orientation control.

• Control with Obstacle Avoidance:

  • Developing control systems to maintain stability, avoid obstacles, and transition between modes ensures safe and efficient operation.

• Multi-Agent Architecture for AUVs:

  • Task Switching and Priority-Based Allocation: Enable each AUV to take on roles based on its capabilities, allowing for dynamic task switching and reallocation based on mission priorities. This will maximize resource utilization and ensure that high-priority tasks receive attention from the most capable agents.
  • Leader-Follower Control Strategy: Implementing a leader-follower setup will create efficient coordination in complex environments, ensuring a lead AUV can guide the group while followers maintain formation. This approach enhances mission cohesion and simplifies multi-agent movement through challenging areas.
  • Inter-Agent Communication: A reliable communication system among AUVs is essential for sharing real-time data on position, task status, and environmental changes. This will support adaptive decision-making, allowing agents to respond collectively to evolving mission conditions.
  • Obstacle Avoidance and Navigation: Equipping AUVs with autonomous obstacle detection and avoidance ensures safe operation and efficient navigation. This capability will be critical for uninterrupted multi-agent operations in dynamic underwater environments.

• Hardware Development and Deployment:

  • Building and integrating reliable hardware that can withstand dual-environment operations is essential, especially given the added challenge of balancing the combined vehicle-drone system for optimal aerial and aquatic performance.

Implementation Scheme • Modeling: Begin by developing accurate mathematical models for both surface and underwater dynamics of your vehicles. • Simulation: Utilize simulation environments (like MATLAB/Simulink, Gazebo, or ROS) to test control strategies in virtual scenarios before deploying on actual vehicles. • Hardware Integration: Once simulations yield satisfactory results, implement the control strategies on the physical vehicles, ensuring to integrate with sensors and actuators for real-time feedback. • Testing and Validation: Conduct extensive testing in controlled environments, gradually introducing complexity (e.g., obstacles, varying terrains) to validate the robustness of the control strategies.

Video

# ROS2 Jazzy Multi-Agent Simulation Project

This repository contains two main ROS2 packages:

- `message_type`: Contains custom ROS2 message types.
- `surface`: Contains launch files, nodes, and logic for simulation, control, and decision-making.

---

## 📦 Packages Overview

###`message_type`
Defines all custom message types used throughout the project.

###`surface`
The core package for launching Gazebo simulations and running multi-agent systems with obstacle avoidance and decision making.

---

🔧 ## Installation & Build Instructions

1. Clone the packages into your ROS2 Jazzy workspace:

   ```bash
   cd ~/your_ros2_ws/src
   git clone <repo_url_for_message_type>
   git clone <repo_url_for_surface>

2. Build the workspace:

   cd ~/your_ros2_ws
   colcon build

3. Source the workspace:

   source install/setup.bash

---

🚀 Usage

🏞️ Launch the Gazebo World

ros2 launch surface comb_new_launch.py

🧭 Run Control & Planning Nodes

• Start ROAM (Rotational Obstacle Avoidance Method):

  ros2 run surface under_study

• Start DMM (Dynamic Modulation Matrix Method for Obstacle Avoidance):

  ros2 run surface new_control

🤖 Multi-Agent Simulation

• Basic Multi-Agent Launch:

  ros2 launch surface multi_agent_launch.py

• Multi-Agent Navigation & Decision Making:

  ros2 launch surface multi_agents_launch.py

📤 Task Publisher

ros2 run surface task_publisher

---

📝 Notes

• This project uses ROS2 Jazzy.
• Make sure Gazebo and all necessary ROS2 dependencies are properly installed and sourced.

---