Subject:
|
Re: maze solving algorithm
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Fri, 20 Jun 2003 20:57:14 GMT
|
Original-From:
|
scott davis <[rcx2man@hotmail]Spamcake[.com]>
|
Viewed:
|
965 times
|
| |
|
|
> Scott wrote:
>
> > The robot was supposed to flood the maze starting in the
> > finish square mapping how far each square is from the finish square.
> > There is still some problems with the program ...
>
> So you used the flood-fill or bellman algorithm described here:
>
> http://micromouse.cannock.ac.uk/maze/solving.htm
It is a variation if it. Using NQC I was unable to create a second array.
> (Jona showed that URL in his posting at 2003 may 12).
>
> A synchro drive robot allways points in the same direction, so he could be
> controlled via PC by IR tower.
>
> I see problems arising from the fact, the robot dont "run on rails". The
> question is, how the robot centers in the cell. As i understand Scotts • design,
> you use rotation sensors (one or two?) to measure the way to the next • cell. Or
> is this wrong, how does the robot know that he has reached the next cell? • Will
> the robot use its distance sensor to center in the cell? If the robot does • not
> center in the cell, positioning errors will add and the robot will touch • the
> walls or will stop in between cells, where he cannot determine the walls • around
> the cell.
>
> Is this one of the problems? Can you tell us something about this stuff?
The robot had one rotation sensro that monitored how far forward/backwards
it had traveled. This was to try to keep the robot in the center of the
maze. Which worked reasonably well
.
The only problem with the robot is when it turned its wheels to go left or
right, it would "dance" a little.
I built stops to keep the turntables from turning more than 90 degrees.
which way the motors were directed determined which way the robot moved.
rev made the robot go left
fwd made the robot go right
The major problems were in the programing.
The biggest one being how to get from one square to a unvisited or frontier
square.
> Ralph
Scott
|
|
Message is in Reply To:
 | | Re: maze solving algorithm
|
| (...) So you used the flood-fill or bellman algorithm described here: (URL) showed that URL in his posting at 2003 may 12). A synchro drive robot allways points in the same direction, so he could be controlled via PC by IR tower. I see problems (...) (21 years ago, 20-Jun-03, to lugnet.robotics)
|
2 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
