Re: [TuxKart-devel] Ideas for AI

[Steve B. <sjb...@ai...>]

Charles Goodwin wrote:

> ...so we need a new scheme. ;)

The thing I'd like to consider is something like a 'vector field'.

A large array (and as I explained earlier - it would be LOTS of memory)
where for every (say) 2meter x 2meter square there would be (say) two bytes.

Packed into these 16 bits there would be:

1) The compass direction to steer along in order to reach the best
    racing line in the optimal way from this point.

2) The distance to the nearest powerup that's further down the
    track and 'reasonably' gettable without deviating wildly from
    the racing line.

3) The direction to that powerup.

Armed with this information, the AI can figure out where to go if they
want a powerup, where to go if they want to just drive - and (using the
distance metric and the difference between the two directions) whether
they really want to go out of their way to collect said powerup.

Now - there are some complications.

The map changes when a powerup is collected.  This change is far-reaching
and would result in a ton of complicated realtime calculations.

The good direction is sometimes blocked by another kart.

There is still a problem with bridges, crossovers, and dynamic things.

---------------------------- Steve Baker -------------------------
HomeEmail: <sjb...@ai...>    WorkEmail: <sj...@li...>
HomePage : http://www.sjbaker.org
Projects : http://plib.sf.net    http://tuxaqfh.sf.net
            http://tuxkart.sf.net http://prettypoly.sf.net
-----BEGIN GEEK CODE BLOCK-----
GCS d-- s:+ a+ C++++$ UL+++$ P--- L++++$ E--- W+++ N o+ K? w--- !O M-
V-- PS++ PE- Y-- PGP-- t+ 5 X R+++ tv b++ DI++ D G+ e++ h--(-) r+++ y++++
-----END GEEK CODE BLOCK-----