Journals:Swarm


04/24/11: Journal Entry 8, Week 4/18-4/24

Images created for our simulation:
Bee.gifBug.gifCritter.gif

Tasks accomplished:

  • Today we did a lot of debugging, but we are still getting errors from our grid, tomorrow we are scheduled to meet with someone who we think can help
  • We discussed possible methods for combining our algorithms

Next Steps:

  • We still have a lot of debugging to do
  • We need to complete our presentation for Thursday
  • We need to figure out what an appropriate graphic representation of a victim would be

04/24/11: Journal Entry 7, Week 4/18-4/24

Tasks accomplished:

  • This week we have done a LOT of coding
  • We have built classes/objects for a bee, cockroach, ant, and victim, but the codes are not working the way we would expect in our simulations

Next Steps:

  • We have a lot of debugging to do
  • We still haven't found a good way to combine the algorithms
  • We need to figure out what an appropriate graphic representation of a victim would be

04/17/11: Journal Entry 6, Week 4/12-4/17

Tasks accomplished:

  • Further explored the pros and cons of different algorithms for bee colony and ant. Narrowing options to begin implementation and evaluate feasibility of algorithm in a GridWorld context.
  • Started consolidating ideas on how to combine bee and ant algorithms based on the division of labor in the bee colony system with the ant optimization as a method of travel.

Next Steps:

  • Get further help in coding by using online tutorials and talking to TA's and professors in the CIS department
  • Code and implement all algorithms within the context of classes that work with the grid system of the CollegeBoard's GirdWorld
  • Begin working together on combining aspects of different algorithms

04/17/11: Journal Entry 5, Week 4/12-4/17

Tasks accomplished:

  • Read up APCS CollegeBoard documentation of GridWorld
  • Modified Cockroach Algorithm to move according to the environment of GridWorld
  • Decided that the local optimum are candidate locations for danger zones. Thus, swarming to danger zones, gather information and can classify if it is indeed a danger zone.
  • Started consolidating ideas on how to combine algorithms. Discussing creating a mini-map where robots map danger zones onto an existing map of the area.

Next Steps:

  • Get further help in coding by using online tutorials and talking to TA's and professors in the CIS department
  • Code and implement all algorithms within the context of classes that work with the grid system of the CollegeBoard's GirdWorld
  • Begin working together on combining aspects of different algorithms

04/11/11: Journal Entry 4, Week 4/4-4/11

Tasks accomplished:

  • Completed C6
  • Re-structured plan moving forward
  • Went to CIS tutoring center office hours on Sunday night but did not get adequate help

Next Steps:

  • Get further help in coding by using online tutorials and talking to TA's and professors in the CIS department
  • Code and implement all algorithms within the context of classes that work with the grid system of the CollegeBoard's GirdWorld


04/09/11: Journal Entry 3, Week 4/4-4/11

Tasks accomplished:

  • Decided the model an "urban" search and rescue model

Next Steps:

  • Go to CIS office hours to better understand installation of the GridWorld system

A description of our extension of the object oriented GridWorld system is bellow:

We will use base code from the CollegeBoard's GirdWorld simulation. This system has in place "bugs" "rocks" "flowers" "critters" and other actors that interact in a "grid" class. Instead of using the default actors, we will create "ant," "bee," and "cockroach" actors, which will implement implement the inspect specific algorithm, as well as a "mix" actor, which will use elements of each. We will alter the "grid" to be our simulation and fill it with "brick" actors as insurmountable objects, "rubble" in which "actors" lose their ability to see and sense others, "water" in which actors slow down, and "victim" which is the ultimate target. If a "victim" is under a "brick" then multiple ants, bees, cockroaches, or mix actors will be needed to extract the "victim." "Victim" "health" will decrease over time, making fast rescue an important factor in success. The algorithms will be set as methods of each actor class, and will interact with each other through the grid. For example a cockroach method might scan whatever part of the grid is in "eye sight" (areas within a certain radius that are not blocked by other objects), and move towards the next closest cockroach.


04/03/11: Journal Entry 2, Week 3/28-4/3

Tasks accomplished:

  • Begun research on creating simulation environments. Different S&R situations have very different needs and limitations.
    • For example an urban environment ruble or flooding might be the largest obstacles, but a rural environment might have plant or animal life, fallen trees, temperature limitations, weather restrictions, etc.
  • Found that Java GridWorld from the AP Computer Science curriculum might be a good jumping off place
  • Reviewed various algorithms inspired by bees, and pseudo-coded a bee-inspired algorithm, specifically finding that the following sites might work best for our purposes:

Next Steps:

  • Code in Java the simulation environment first, and then begin coding for insect inspired search and rescue so that all insect codes are compatible with the enviornment


04/01/11: Journal Entry 1, Week 3/28-4/3

Tasks Accomplished:

  • Experimented with Java and Eclipse IDE.
  • Ran through short tutorial on Java Programming Language.
  • Re-read Chen et al. paper titled Cockroach Swarm Optimization.
  • Further researched Cockroach Swarm Optimization.
  • Created pseudo-code for Cockroach Swarm behavior and a Cockroach object.


03/01/11: Sample Entry 2

Upcoming goals:

  1. Begin Project!


03/01/11: Sample Entry 1

Tasks Accomplished:

  1. Wiki Page for C.3
  2. Presentation for C.4
  3. Created journal!