Progress for week 20 (2015)

Fra Robin

(Forskjeller mellom versjoner)
Gå til: navigasjon, søk
(André)
(Stian)
 
Linje 42: Linje 42:
=== Accounting ===
=== Accounting ===
 +
* Looked at evolution of pareto front over time for both training and test.[https://youtu.be/6RkksoWAaUI]
 +
* Found a better pareto front calculation utility for Matlab. Optimized it further by dividing the search space.
 +
* Looked at several timeout methods and how the hypervolume is affected.

Nåværende revisjon fra 2. jun 2015 kl. 08:56

Innhold

Student template (copy this for your entry)

Budget

  • Todo 1
  • Todo 2

Accounting

  • Done 1
  • Done 2

Eivind

Budget

  • Obtain global position.
  • Lasercut new top level.
  • Finish power-system, including start-functionality.
  • Continue tweaking communication between robot and AI.
  • Print last remaining parts.

Accounting

  • Modified existing top level to fit instead.
  • Powersystem finished (excluding start-functionality).
  • Communication between control-system and AI is working well.
  • All parts have been printed.
  • Set up system for detecting opponent.

André

Budget

  • Put the tower on top of the robot, and 3D print beacon 3.
  • Finish the server, and implementation between the sensors, positioning system, and all the other system
  • Write the final program for the sensor in different sones
  • Make some tests with the positioning tower, and try tweaking it to be more precise
  • Tweak the values of the kalmanfilter to make it optimal

Accounting

  • Done with the competition

Stian

Budget

  • Write
  • Fix the pareto front Matlab code. Try to reduce the search space.
  • Look at results.

Accounting

  • Looked at evolution of pareto front over time for both training and test.[1]
  • Found a better pareto front calculation utility for Matlab. Optimized it further by dividing the search space.
  • Looked at several timeout methods and how the hypervolume is affected.
Personlige verktøy