This repository contains the notebook "Formation Control with Leader", a detailed analysis and implementation of multi-agent systems based on potentials. The notebook guides through the development of a formation control system for a leader and a group of agents that maintain a circular arrangement around the leader.
The project simulates a multi-agent system that includes:
- 🤖 Leader: A guiding agent that moves toward a goal following an attractive potential.
- 🌀 Agents: A group of N agents that maintain a circular formation around the leader using a formation potential.
- 🚩 Leader Control: Directing the leader toward a fixed goal.
- ⚙️ Formation Maintenance: Ensuring that non-leader agents respect the desired arrangement.
- 🔒 Robustness: Testing the stability of the formation even with a moving goal (at a constant speed).
- 📊 Performance Analysis: Calculating the maximum goal velocity that allows the formation to remain intact.
- Attractive Potential: Directs the leader toward the goal.
- Formation Potential: Forces agents to maintain formation relative to the leader.
🔸 Main Notebook:
Contains Python code with step-by-step explanations. It uses the following libraries:
- 🧮
numpyfor numerical computations. - 📈
matplotlibfor visualization.
🔸 Parameter Tuning:
Demonstrates how to optimize potential parameters to ensure robustness and stability.
Since the plots in the notebook are static, GIFs are generated during execution using matplotlib.animation, allowing for a dynamic visualization of agent movements.
This project was developed by Giovanni Stefanini as part of an analysis on multi-agent formation control for the Multiagent Systems exam.