[Lua] Missle Avoidance/Prediction (Naval/Aerial Nuking)

Come in here to talk about your sky-net style world-destroying super bots!

Moderator: Defcon moderators

User avatar
Ace Rimmer
level5
level5
Posts: 10803
Joined: Thu Dec 07, 2006 9:46 pm
Location: The Multiverse

[Lua] Missle Avoidance/Prediction (Naval/Aerial Nuking)

Postby Ace Rimmer » Fri May 07, 2010 4:16 pm

Here's the big question; once you know an object's (missile) 'path' (let's say from first seen to current in increments of one, e.g. start 1 Long, 1 Lat; 2,2; 3,3; 4,2; 5,1 current 6,0), how in the world do you figure out where it's going to land?

I'm sure there's some sophisticated math based on the map in Defcon being 'close' to cylindrical, right?

Why would a person want to know this? Simple; to avoid having ships/planes being destroyed with nukes and also to perform emergency evacs of carriers/airbases.

Now, at this point I've just started using unitID's as tables to keep track of data and haven't yet come up with a way to log the path of any given missile, but I believe that will just be a matter of time. My Bot has already been the victim of self nuking in combat, so even (way in the future) if a Bot is able to work in a team, it would need to be able to avoid allied nukes.

Help?
Smoke me a kipper, I'll be back for breakfast...
Montyphy
level5
level5
Posts: 6747
Joined: Tue Apr 19, 2005 2:28 pm
Location: Bristol, England

Postby Montyphy » Fri May 07, 2010 4:30 pm

You would possibly need to gather some start, mid and end coords for nukes launched from various points on the globe. From there you could reverse engineer the curve Defcon uses.
Last edited by Montyphy on Fri May 07, 2010 4:50 pm, edited 1 time in total.
Uplink help: Check out the Guide or FAQ.
Latest Uplink patch is v1.55.
User avatar
bert_the_turtle
level5
level5
Posts: 4795
Joined: Fri Oct 13, 2006 6:11 pm
Location: Cologne
Contact:

Postby bert_the_turtle » Fri May 07, 2010 4:40 pm

Hmm, I don't think anyone has worked out the precise formula for the trajectories yet. For starters, you can extrapolate the path using a cubic polynomial fit (or, possibly more accurate, a circular line) That won't give you the point they'll land on, though. Even if the path formula was completely reversible, I'd consider it cheating if a bot used that to extrapolate the exact nuke landing spot.
User avatar
Ace Rimmer
level5
level5
Posts: 10803
Joined: Thu Dec 07, 2006 9:46 pm
Location: The Multiverse

Postby Ace Rimmer » Fri May 07, 2010 4:43 pm

bert_the_turtle wrote: Even if the path formula was completely reversible, I'd consider it cheating if a bot used that to extrapolate the exact nuke landing spot.

Why? Players do it in the game all the time, just via eyesight and experience.
Smoke me a kipper, I'll be back for breakfast...
martin
level5
level5
Posts: 3210
Joined: Fri Nov 19, 2004 8:37 pm

Postby martin » Fri May 07, 2010 5:38 pm

I have to agree with Ace here. The advantage bots have is their absolute mathematical precision and speed.
GENERATION 22:The first time you see this, copy it into your sig on any forum and add 1 to the generation. Social experiment.
User avatar
bert_the_turtle
level5
level5
Posts: 4795
Joined: Fri Oct 13, 2006 6:11 pm
Location: Cologne
Contact:

Postby bert_the_turtle » Fri May 07, 2010 9:50 pm

Ace Rimmer wrote:
bert_the_turtle wrote: Even if the path formula was completely reversible, I'd consider it cheating if a bot used that to extrapolate the exact nuke landing spot.

Why? Players do it in the game all the time, just via eyesight and experience.
Really? The exact landing spot? I'm sure you can extrapolate the trajectory, but can you really look at the shape of it and tell when the nuke will detonate?
User avatar
Ace Rimmer
level5
level5
Posts: 10803
Joined: Thu Dec 07, 2006 9:46 pm
Location: The Multiverse

Postby Ace Rimmer » Fri May 07, 2010 9:55 pm

bert_the_turtle wrote:
Ace Rimmer wrote:
bert_the_turtle wrote: Even if the path formula was completely reversible, I'd consider it cheating if a bot used that to extrapolate the exact nuke landing spot.

Why? Players do it in the game all the time, just via eyesight and experience.
Really? The exact landing spot? I'm sure you can extrapolate the trajectory, but can you really look at the shape of it and tell when the nuke will detonate?

Perhaps not the exact spot every time, but I would guess if I kept it on real time and studied each nuke (annoying other players of course), I could get a pretty high average.

Either way, I don't expect the bot to be able to predict the exact spot.
Smoke me a kipper, I'll be back for breakfast...
martin
level5
level5
Posts: 3210
Joined: Fri Nov 19, 2004 8:37 pm

Postby martin » Sat May 08, 2010 12:08 am

Oh I see what you mean, well as far as I know there is no way for the bot to know the exact spot, just extrapolate its trajectory and guess when it will detonate along that path
GENERATION 22:The first time you see this, copy it into your sig on any forum and add 1 to the generation. Social experiment.
User avatar
poptart911
level1
level1
Posts: 33
Joined: Thu Jan 13, 2011 6:04 pm
Location: Iowa, US
Contact:

Postby poptart911 » Sat Jan 15, 2011 3:45 am

martin wrote:Oh I see what you mean, well as far as I know there is no way for the bot to know the exact spot, just extrapolate its trajectory and guess when it will detonate along that path


Right, the trajectory could include multiple targets. There's no way to know which one would be the end.
User avatar
trickser
level5
level5
Posts: 1826
Joined: Thu Mar 06, 2008 2:15 pm
Location: The Senate ; GMT+1
Contact:

Postby trickser » Thu Feb 03, 2011 5:25 pm

The curving of a nukes trajectory is ihmo a function of its relative position to its start/target distance (or something like that), just like its speed. With knowing the exact formula you can extrapolate pretty much the exact target. I made an attempt figuring that formula a while back. For the special case of common nuke curves ( :P ) its should already be good enough, if you know 2 or 3 nuke positions.

Return to “AI Bots”

Who is online

Users browsing this forum: No registered users and 5 guests