2010Fall Project 2

From Autonomous Multirobot Systems

Jump to: navigation, search

Contents

Heterogeneous Soccer Team

Overview

This project carries forward the last project in ASCII soccer. The task for this assignment is to investigate the impact of relaxing the requirement for a homogeneous solution.

Assessment and Rules

Your code will be assessed by playing a game to 1000 points against YOUR PREVIOUS TEAM.

Do: Create individual playerX() functions. Each player should have its own function. If you chose, you may elect to have each robot be identical to the others, but this is no longer required.

Important constraint: Your player routines may contain at most 700 characters (not counting comments). You may use C macros and defines to reduce the size of your code, but the text you use to encode them counts towards your 700 character budget.

What to Turn In

You should submit a short report and well documented code via T-Square. We will expect two files:

  • main.c, your code and
  • report.pdf

Please do not submit additional files or word (.doc) documents! The report can be very brief (1 page), and it should contain the following sections.

  • Your name.
  • A description of your algorithm. Is it diverse? In what way? (note it is OK if it is not diverse).
  • Compute the social entropy of your team using the metrics discussed in class.
  • Compute the behavioral difference (approximation is OK) between each of the players on your team using the metrics presented in class. The result should be a 4 by 4 table. If your team is homogeneous... I'll think of another thing for you to do here :-).
  • An assessment of the situations where you think the program will work well, versus those where it does not work well.
  • For evaluation purposes, all of your code should be contained within main.c.
  • Does your algorithm beat your previous team in a game to 1000 points?

Our assessment of your of writeup will center on these questions: Do you describe your algorithm sufficiently well that someone else could implement it? Do you effectively analyze your algorithm, explaining its strengths and weaknesses?

Personal tools