Staredit Network > Forums > Null > Topic: Laplace Transformations
Laplace Transformations
Mar 25 2009, 5:28 am
By: A_of-s_t  

Mar 25 2009, 5:28 am A_of-s_t Post #1

aka idmontie

Well, I'm bored in my mathematics classes, so I was browsing around my old text books I bought and I came across Laplace Transformations and I was wondering what they are actually used for.

F(s) = L{f(t)} ; f(t) = L^(-1){F(s)}

Is this a form of extending the domain of a function who takes real values over the complex domain? And for the inverse, does it take a function whose domain takes complex values and transforms it into the real domain? Or am I all mixed up about Laplace Transformations?

(And no, linking to wikipedia won't really help in this situation)



Personal GitHub
Starcraft GitHub Organization - Feel free to request member status!
TwitchTV

Mar 25 2009, 6:29 am Vrael Post #2



The Laplace Transformation is the integral from 0 to infinity of ( e^-(s)t )( f(t) ), which as you mentioned gives you a function F(s)

Take the simplest example where f(t) = 1
I will use the notation "0|infinity" as "the integral from 0 to infinity"

L{1} = 0|infinity e^-st dt
=lim_A->infinity [ -(1/s)e^-s(A) - (-((1/s)e^-s(0)) ]
= 1/s

So for f(t) = 1, F(s) = L{1} = 1/s
(likewise L{n} = n/s)

The next simplest example is L{e^at}
= 0|infinity ( (e^-st)(e^at) )dt
= 0|infinity (e^-(s-a)t)dt
which evaluates to
F(s) = 1/(s-a)

For numerous functions f(t) we can build a table of F(s)'s, much like the table of integrals in the back of a calculus book.
If we have the function in some F(s) form, we can use L^-1{F(s)} to get f(t)

These operations become very useful in the analysis of differential equations

Take y'' - 2y' + y = cosine(t)

If we take the Laplace transform of the function it becomes (and Laplace transforms are linear so you can move the two out of the curly brackets)
L{y''} + 2L{y'} + L{y} = L{cos(t)}

Another useful property of the L{ } is that L{y'} is directly related to L{y} ( didn't derive it here, but these are the equ's)
L{y'} = sL{y} - y(0)
L{y''} = (s^2)L{y} - sy(0) - y'(0)

If we take the initial condition of y and y prime to be 0 to simplify things, and we substitute in for L{y'} and L{y''} we get
(s^2)L{y} + 2sL{y} + L{y} = L{cos(t)}

From a table of F(s)'s we know that L{cos(t)} = (s^2)/[(s^2) + (1^2)] (I write 1^2 because the coefficient of the t in cos(t) is 1)

so we have
(s^2)L{y} + 2sL{y} + L{y} = (s^2)/[(s^2) + (1^2)]

(s^2 + 2s +1)L{y} = (s^2)/[(s^2) + (1^2)]

(s+1)(s+1)L{y} = (s^2)/[(s^2) + (1^2)]

L{y} = (s^2)/[(s^2) + (1^2)][(s+1)(s+1)]

If we use partial fractions to separate the big 's' expression into smaller parts, we end up with expressions like 1/(s-a), which we know from our table is the L^-1{e^at}. After solving the L^-1{y} on both sides, we are left with y, which is the solution to our differential equation.



None.

Options
  Back to forum
Please log in to reply to this topic or to report it.
Members in this topic: None.
[10:31 am]
NudeRaider -- blessed innocent soul knows nothing of the strife we had before EUDs were discovered :teehee:
[2026-6-23. : 3:29 am]
DarkenedFantasies -- Probably just didn't care. For example, at some point before release, they've updated the graphics of some of the Protoss buildings (Forge, CyberCore, Citadel, Observatory, Arbiter Tribunal), but instead of properly re-rendering them with edited 3D models, they did crappy copy-paste jobs on the rendered graphics.
[2026-6-22. : 8:35 pm]
Ultraviolet -- :wob:
[2026-6-21. : 11:38 pm]
Symmetry -- :wob:
[2026-6-21. : 4:56 am]
Ultraviolet -- I suppose we'll likely never know, but my guess would be that they already saw it operating successfully and there was no monetary incentive to finish the original work. And the dev cycle in old school Blizzard was so hectic, it's possible it just got forgotten about after the original game got released. Plus there's an element of existing MPQ files that were packaged with the original discs becoming outdated if they updated it. And it's not like they remade the original MPQs, they just made new ones for BW specifically
[2026-6-21. : 4:26 am]
Oh_Man -- so that makes me think maybe the theory they are unfinished is not true and its a deliberate design decision, coz why not finish them wen ur making brood war?
[2026-6-21. : 4:25 am]
Oh_Man -- the thing is thos buildings are from classic. that means they went ahead and made brood war without ever finishing the 'unfinished' buildings
[2026-6-20. : 6:15 pm]
Ultraviolet -- Yeah he's talked about a lot of that stuff in his casts before. It seems plausible. Especially knowing how Blizzard of yesteryear operated.
[2026-6-20. : 3:47 pm]
NudeRaider -- to clarify: couldn't recall the behavior for every single Protoss building but I was aware the disparity exists.
[2026-6-20. : 3:43 pm]
NudeRaider -- Contained nothing new for me. Didn't know all building's behavior, but very much all unit's. Also Terran balance whine - also nothing new :lol:
Please log in to shout.


Members Online: Oh_Man