Wednesday, June 5, 2013

More Math

So today I finally (I know, it's only been a week and half, and already it's "finally") got to code something! (In Matlab, which is like well, kind of code. And indexes from one.. who does stuff like that?)

I didn't really know Matlab. like at all.* So it was fun. So was finding all the different ways Germans and Americans describe math. Perhaps I'm not math-literate enough, but I wasn't aware of any numerical mathematics branch. Also turns out we Americans are lazy. You thought Newton method was enough. NO! It's Newton-Rhapson Rhapsid? Rhapsody? Whatever.. support the stereotype and what not.                               *well fine, I used a bit for linear algebra assignments. but it doesn't count

There was also that time when everything's a variable and you don't remember which one your solving for so you pick one and it's the wrong one so all your matrices are singular (or approaching singularities). And that time where the spherical joints were, then weren't, then were again spherical. (don't ask)

And math. Math everywhere. I'm pretty sure I'll be rotating sheep in my dreams. So if the sheep jump in an arch over the fence (that's what my sheep do while I count them.. I mean having them stand in nice grids eating grass would make it waaay too easy), what are the coordinates of sheep n with relation to sheep n+1 at time step n.... yea I'll fall asleep.

However, on the bright side, I do sort of know Matlab programming now. (quite a marketable skill I'm told) and I can convert from target points/orientations to piston lengths and back the other way! It's a miracle (practically).

Right. So the project. We have this hexapod. It does not have six legs. or 5 or 4 or any. well. eh. It's a base platform with 6 pistons going towards the top platform. and because of cool kinematics stuff, it's a sturdy, strong, fast way to move about to points. Unfortunately doing stuff like that (because of not so cool kinematics) is hard. Which is the point (If I understand it correctly) of this research.

Total Product of today: Pretty pictures (that don't show anything useful, but are proof semi-definite(ly) that I have something that will work once I fix the whole joints that are but aren't but are spherical. and the size of triangles. (That's more of a making graphs look pretty thing though))

No comments:

Post a Comment