Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Walking Equation Help

  1. Apr 7, 2005 #1
    I am new here and i am a java programer and i needed a physics equation. I need to get an equation that can allow a robot to calculate how to get from point a to b the best way. I have been thinking of something like that but keep drawing blanks. Any help
  2. jcsd
  3. Apr 7, 2005 #2
    You are not giving use much to go on. Is it a straight line? does the robot have a map or does it need to observe it surroundings?

    Graph theory has many common algorithms to determine the lowest cost path. Is this what you are looking for?
  4. Apr 8, 2005 #3
    It is going in a straigh path the distance is going to be like 50 feet or something.
  5. Apr 8, 2005 #4


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    .. and the simple equation of a straight line like y = mx + b just doesn't cut it?

  6. Apr 8, 2005 #5


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    If the co-ordinates of the 2 points are A(x1,y1) and B(x2,y2), the equation of the line passing between them is :

    [tex]y - y_1 = \frac {y_1 - y_2}{x_1 - x_2} ~ \cdot ~ x - x_1 [/tex]
  7. Apr 8, 2005 #6
    Please tell me there is something more too it then this, there are bumps on the ground, there is a hill in the way, or some sort of obstacles. It is not flat ground for 50 feet is it?
  8. Apr 8, 2005 #7


    User Avatar
    Science Advisor
    Homework Helper

    Another question:is this a geodesic or a brahistochrone problem...?

  9. Apr 8, 2005 #8


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    I vote for the speed bump....

  10. Apr 14, 2005 #9
    It is a straight path for 50 feet. The equation is meant to find figure out how big the steps have to be so it doesn't fall forward or something other than that there is not going to be a sudden unexcpected force accting on the robot.
  11. Apr 14, 2005 #10
    That's too ambitious kura_kai to be handled by a simple Java program. There are two things you can consider to drastically improve you robot design: 1. use the reactive agent architecture instead of symbolic reasoning (or at least hybrid), 2. Incorporate some form of learning to your robot, so that the latter itself decides which are the most optimal movements and step sizes. Sth like reinforcement learning.
    Good luck!
  12. Apr 14, 2005 #11
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?

Similar Discussions: Walking Equation Help
  1. Equation help (Replies: 2)

  2. Blasius equation help (Replies: 5)