razzler

I'm doing a line maze robot without the use of a microcontroller. I use backtracking to determined the correct paths and store them in memory but the thing is, after the robot reaches the end goal, it has to go back and traverse the successful path back to the start line.My instructor advises us to follow a stacking algorithm but how do we determine that the data inputted is that of the successful path??


please help...

Torben

When you say "without the use of a microcontroller", do you mean your control software is running on a PC or similar?

A stack should do the trick. Think about what happens as the robot moves through the maze: as you explore each new segment, push it onto the stack; if it's a dead end, backtrack to the previous unexplored segment branch and pop the bad segment off the stack. What's left when you finally make it to the end of the maze is your path.


Torben

razzler

i'm using a combination of common ICs that would be some comparators and counters and memory.

btw, thanks. it's much clearer now. ^_^

Torben

Glad to have helped. Pretty impressive doing this without a uC or PC. We get a lot of cries for help on this board from "final year engineering" students whose schools (or instructors) don't seem to teach them anything.

Good luck with the project!


Torben

Krumlink

To store data like that it will be near impossible, unless you have a ton of space for transistors. Stick to MC.

__________________
MechTronics

Torben

. . .except he said he's using memory ICs, so as long as he's got the support circuits and clocking down he should be good to go.


Torben

Krumlink

Oh, I diddnt see that. I thought he wanted to just use comparators

It would still be easier to use a Microcontroller.

What exactly is the maze supposed to be/solved?

__________________
MechTronics

razzler

yup...the instructor expects the maze to be solved by the robot...^_^;;;

any more ideas?it'd be great help ^_^

Nigel Goodwin

I don't see as you're going to 'solve' a maze in hardware?, it would be far too complicated.

However, you could get your robot to the centre by simply following the left hand wall - but that isn't really 'solving' the maze.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

razzler

the block diagram is like this


robot sensors->control logic->counter->memory

we'll be using a 4x1 MUX as our control logic and maybe a binary counter


what do you think?

Pommie

Sounds like a near impossible task to me. You would end up with something like the early pong games. A board about 20 inches on each side fully populated with ttl chips.

Mike.

razzler

is it?how so?

Nigel Goodwin

Because you need a computer - so you're going to need to build one from scratch!.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

razzler

so our instructor gave something impossible to do so I'd most likely fail

Nigel Goodwin

Depends what you mean by 'solve'?

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk