Multi-Algorithm UAV Path Planning and Simulation

This project demonstrates a dynamic UAV (Unmanned Aerial Vehicle) simulation where a leader UAV moves in a 2D environment with circular obstacles, and multiple followers track the leader using different path planning or tracking algorithms (A*, Predictive, Direct).

Installation

1. Clone or Download the repository:

   git clone https://github.com/satilmiskabasakal0/UAV-Path-Simulation.git
   cd UAV-Path-Simulation
   

Usage

Install Dependencies:

pip install -r requirements.txt

Configure Simulation

Open config.yaml to change any parameters (like area_size, sim_time, dt, etc.).

Run Simulation

Simply run:

python main.py

This launches a Matplotlib animation window to display the leader and multiple followers in action. At the end (or while closing), it attempts to save an animation in animations/ as animation_x.mp4.

Add or Modify Algorithms

tracking_algorithms.py contains the logic (A*, Predictive, Direct). Implement your own or tweak existing ones. Adjust followers in main.py (e.g., different speeds, different modes).

Key Features

1. Dynamic Configuration via config.yaml:

   • The simulation parameters (e.g., area_size, simulation_time, dt, number of obstacles, etc.) can be easily changed in config.yaml.
   • No code modification is needed to change the simulation scale or obstacle counts—just update the YAML file.

2. UAV Definitions in uav.py:

   • Leader (Leader class):
     • Moves randomly, with obstacle avoidance.
     • Changes direction at configurable intervals (see change_interval_time in config.yaml).
     • If inside an obstacle, it finds an exit direction and moves out.
   • Follower (Follower class):
     • Follows the leader using various algorithms.
       • A* (A-Star) path planning
       • Predictive (estimates leader’s future position)
       • Direct (moves directly toward leader’s current position)
     • Avoids obstacles in the same manner as the leader.

3. Obstacle Generation via obstacles.py:

   • Creates circular obstacles placed randomly, with radii also randomized.
   • Ensures minimal overlap between obstacles.
   • Checks whether a UAV is inside an obstacle and provides a direction to exit.

4. Dynamic Path Planning:

   • For each time step, the leader updates its position.
   • Each follower picks the relevant tracking algorithm and updates accordingly.
   • Follower path or direction is re-planned as needed.

5. Visualization and Animation:

   • Uses Matplotlib to display:
     • The simulation area,
     • Obstacles (in red),
     • Leader’s path and marker,
     • Each Follower’s path and marker (in different colors).
   • Saves the animation to MP4 if FFmpeg is installed.

6. Easily Extendable:

   • Add new algorithms in tracking_algorithms.py.
   • Add or remove followers in main.py.
   • Adjust speeds, intervals, obstacle counts, and more in config.yaml.

Project Structure

UAV-Path-Simulation/
├── main.py                # Main entry point for running the simulation and animation
├── uav.py                 # UAV classes (Leader, Follower) - obstacle avoidance, updates, etc.
├── obstacles.py           # Creating obstacles, detection (is_inside_obstacle), exit direction
├── tracking_algorithms.py # Implementation of 'astar_follow', 'predictive_follow', 'direct_follow'
├── astar.py               # A* path planning methods (if needed)
├── config.py              # Loads and parses config.yaml for easy simulation changes
├── config.yaml            # YAML file controlling area size, sim time, dt, obstacle count/radii
├── requirements.txt       # Python dependencies
└── animations/            # MP4 animation files saved here