Homework 7

Table of Contents

Task 1 (30 pts)

Read/skim Cooperative transport by ants and robots (PDF) by Kube and Bonabeau (2000) (full citation). Answer the following questions:

  1. What is a suggested solution to the problem of many robots communicating with each other?
  2. What are two problems with robot swarms?
  3. How does an ant decide that “group transport” is needed?
  4. How was the “stagnation” problem solved in experiments with robots?

Task 2 (70 pts)

Modify the Shepherds model in NetLogo:

  • Add a new kind of “breed” called a sheepdog. The sheepdogs will run around dropping “chemical” (scent) which the sheep will avoid.
  • Program the sheepdogs to move, much like the sheep do.
  • Every time a sheepdog moves, it should leave chemical on the “patch” (grid square) that it is currently on.
  • This chemical will diffuse and evaporate. You’ll want to look at the Ants model in NetLogo and copy how the diffusion and evaporation works there.
  • Sheep avoid this chemical (they are scared of the sheepdogs). Again, look at the Ants model; in that model, the ants go towards the chemical. The sheep, on the other hand, should avoid the chemical.

Run some experiments by playing with the number of sheepdogs, diffusion and evaporation rates, etc. Be sure to add sliders to the interface to manipulate these parameters.

Turn in your code (the .nlogo file). If you are not able to transfer it to Linux, email it to me.

Extra credit 1 (10 pts)

Load the Wolf Sheep Predation model. Run a BehaviorSpace experiment (Tools menu, BehaviorSpace). The goal is to find conditions where the populations are stable. Set some parameter ranges and more than one repetition. Measure count wolves and count sheep. Be sure to add a “Stop condition” like count turtles > 5000 to avoid explosions of sheep. Also add a time limit like 3000. Create a spreadsheet of the results. Include this spreadsheet in your homework submission as well as a written summary of what you observed (5-7 sentences).

Extra credit 2 (20 pts)

Create a new strategy for the prisoner’s dilemma model called PD Two Person Iterated. Find the code that starts to custom-strategy and modify it. Experiment with your new strategy and report on how well it works (a few sentences). Provide your new code in your report.

CSE 630 material by Joshua Eckroth is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License. Source code for this website available at GitHub.