Shop OBEX P1 Docs P2 Docs Learn Events
Modelling A Holonomic Wheeled Line Follower — Parallax Forums

Modelling A Holonomic Wheeled Line Follower

HimszyHimszy Posts: 28
edited 2009-10-24 22:53 in Robotics
Hi,

I'm a University student who's been tasked with rather familiar·concept of designing a line following robot. Wanting to go with a "thinking·outside the box" approach and not continue along the traditional 2 motors and caster wheels, I've chosen to go with a holonomic wheel based design.

I've researched the design and have equations and concepts all worked out but I'm looking for a way to be able to model the robot and provide a proof of concept without actually building anything. Which is where I become stuck.... I need some way to be able to model a body with 4 points of contact to the floor, set each wheel at a different speed and examine whether the robot will move as I expect. Can anyone suggest any suitable software please? Being at a well equipped University, I've got lots of programs I can use but rather than learn how to use 10 different ones and fine only 1 is suitable I was hoping someone would be able to steer me into the right direction to begin with.

Regards,
·

Comments

  • allanlane5allanlane5 Posts: 3,815
    edited 2009-10-23 13:48
    I find it's usually better to "think outside the box" once I've solved the problem "inside the box". Meaning if you haven't DONE a 2-motors and caster solution already, that should probably be your first solution. THEN you can try to implement the expansion you're discussing.

    What computer languages are you comfortable with? Because I think you'll have to write your own robot-wheel speed simulation to solve the problem you're talking about.
  • HimszyHimszy Posts: 28
    edited 2009-10-23 14:43
    I already have done it.....just not in the University context (personal projects) hence why I'm keen to go one step further.

    VB (6 and .net), C, C++, Javascript and PHP. Not too sure about writing my own purely as I wouldn't know where to start.
  • W9GFOW9GFO Posts: 4,010
    edited 2009-10-24 03:43
    I bought a WowWee Tri-Bot when it was on Woot! for $30 just so that I can experiment with that kind of drive system. I found the Tri-Bot interesting for about five minutes, it then got very repetitive. It is now completely disassembled. I was glad to find that each motor has an encoder wheel mounted on the shaft - should be able to make an interesting bot out of it.

    Is there a reason you want to have four wheels rather than three? I think three wheels has the advantage that all the wheels will always be in contact with the ground.

    Rich H

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    The Simple Servo Tester, a kit from Gadget Gangster.
  • HimszyHimszy Posts: 28
    edited 2009-10-24 11:45
    I did consider that but I'm keen to make mine slightly different as the implementation and calculations don't vary much and therefore it shows little innovation on my part.
  • W9GFOW9GFO Posts: 4,010
    edited 2009-10-24 22:53
    I don't know how common three wheel versions are compared to four wheeled versions, I know I have seen plenty of both. I think the real challenge will be the control software. I'd like to see them move more fluidly - combining linear motion along with rotation. For instance, make one drive in slow circles around another moving robot while maintaing a constant distance from it and always pointing at the robot which it is orbiting. That would be cool. No one will care how many wheels you use but I think you will have an extra challenge with four wheels unless you have a means to insure that all wheels are always in contact with the surface.

    Rich H

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    The Simple Servo Tester, a kit from Gadget Gangster.
Sign In or Register to comment.