Cell DEVS - Football Player Interaction Model (FPI)

Introduction

This project implements a DEVS-based model for simulating football player interactions, including short passing, long passing, dribbling, and off-ball movement within a defined cell space. The model simulates the dynamic interaction between players and the ball, including state variables such as fatigue, mental state, and action decisions.

Dependencies

This project assumes that you have Cadmium installed in a location accessible by the environment variable $CADMIUM. This dependency would be met by default if you are using the ARSLAB servers. To check, try echo $CADMIUM in the terminal

Build

To build this project, run:

source build_sim.sh

NOTE: Every time you run build_sim.sh, the contents of build/ and bin/ will be replaced.

Running Simulations

Full-Simulation (10x10 Grid)

To test the entire simulation on a 10x10 grid, run:

Without Obstacles

./bin/football_player_interaction config/without_obstacles/10x10_player_config.json

With Obstacles

./bin/football_player_interaction config/with_obstacles/10x10_player_config.json

With Obstacles and Zones

./bin/football_player_interaction config/with_obstacles/with_zones/10x10_player_config.json

With Obstacles, Zones and Roles

./bin/football_player_interaction config/with_obstacles/with_zones/with_roles/10x10_player_config.json

Component Testing (3×3 Grid)

If you are interested in testing specific gameplay components (e.g. short pass, dribble, long pass, off-ball movement), use the 3×3 configuration files:

Files prefixed with 3x3_* in the config/without_obstacles/ and config/with_obstacles/ folders target specific behaviors.

Example

• To test the short passing action on a 3x3 grid:

./bin/football_player_interaction config/without_obstacles/3x3_player_short_pass_config.json

Output Files .csv

Any output generated during the simulation will be saved as a CSV file called grid_log.csv.

To view any output file visually, use the Cell-DEVS Viewer. Simply upload the grid_log.csv file along with the corresponding configuration JSON file that was used to generate it.

Video Files .webm

The recorded simulation videos demonstrate different scenarios using the Cell-DEVS Football Player Interaction Model. Each video corresponds to 10×10 grid gameplay under a specific configuration.

For better clarity, playback speed should be slowed down while viewing the videos.

How to Interpret the Simulations

The Cell-DEVS Viewer displays a 4x3 grid, for the current implementation, where the most important cells to observe are:

• Row 1, Column 1: Represents the player (red when present).
• Row 1, Column 2: Represents the ball (blue when present).
• Row 1, Column 3: Represents the obstacle (green when present).
• Row 2, Column 1: Represents near-obstacle detection (cyan when present).