Difference between revisions of "Gilgalad"
Jump to navigation
Jump to search
(updates) |
(New version of Gilagald) |
||
| Line 4: | Line 4: | ||
| targeting = KNN GF | | targeting = KNN GF | ||
| movement = [[Wave Surfing]] | | movement = [[Wave Surfing]] | ||
| − | | current_version = 1.99. | + | | current_version = 1.99.3 |
| − | | download_link = https://sites.google.com/site/awusa94/robocode/aw.Gilgalad_1.99. | + | | download_link = https://sites.google.com/site/awusa94/robocode/aw.Gilgalad_1.99.3.jar |
| isOneOnOne = yes | | isOneOnOne = yes | ||
}} | }} | ||
| Line 60: | Line 60: | ||
* bug hunting | * bug hunting | ||
* Rewrite some movement code. | * Rewrite some movement code. | ||
| − | |||
* an R-tree is in development. | * an R-tree is in development. | ||
| + | * Use a boosting algorithm for movement | ||
I also plan to try LtL targeting and movement (secret), and Neural networks. | I also plan to try LtL targeting and movement (secret), and Neural networks. | ||
Revision as of 02:16, 9 November 2012
| Gilgalad | |
| Author(s) | AW |
| Extends | AdvancedRobot |
| Targeting | KNN GF |
| Movement | Wave Surfing |
| Current Version | 1.99.3 |
| Download | |
- Sub-pages:
- Version_History - targetingStrategy - movementStrategy
Background Information
- What's special about it?
- My first "good" bot. (Meaning its code is not a huge mess)
- It is the best robot that doesn't use Rednaxela's kD-Tree.
- It is the first robot to use Variable bandwidth in surfing.
- It was one of the first robots to use Precise Positional MEA in targeting.
- How competitive is it?
- As of 9/24/2012 it is ranked 3rd in the 1v1 general rumble.
Strategy
- How does it move?
- it calls setTurnRightRadians() and setAhead(). See Gilgalad/movementStrategy.
- How does it fire?
- it calls setFireBullet(). See Gilgalad/targetingStrategy.
credits
Thanks go to the following people (the list is in no way complete and will probably never be so) :
- Voidious -- Wave Surfing tutorial, and various ideas from Diamond.
- Rednaxela -- Precise intersection and forcing me to improve my kD-tree ;).
Motto
- Strength in numbers.
After adding about five new classification schemes to my movement, I thought this made sense.
Additional Information
- Can I use your code?
- Yes, it is released under the FreeBSD license. (The source is in the jar file and the license should be at the top of each .java file)
- What robots is it based on?
- The code is my own, but I did get ideas from other robots. See the credits for more details.
- What's next for your robot?
- for the immediate future:
- bug hunting
- Rewrite some movement code.
- an R-tree is in development.
- Use a boosting algorithm for movement
I also plan to try LtL targeting and movement (secret), and Neural networks.
