Laplace transform of f(t)=t^n inductive proof?
So we know the laplace transform of this, but could anyone explain how to use an inductive method. Would we suggest letting n=1, solving for that, n=2 solving for that, n=3 solving for that and then assume it for all cases?
setup the integral and see if you can get a recurrence relation
\[F(s)=\int\limits_{0}^{\infty}t ^{n}e ^{-st}dt\]
\[\large \mathcal{L}\{t^n\} ~~=~~ \int\limits_0^{\infty} t^ne^{-st}\,dt\] try by parts
so for \[\int\limits_{}^{}fdg=fg-\int\limits_{}^{}gdf\]
let f=t^n dg=e^(-st)
looks good, keep going..
i'll be using equation mode so bare with me : )
\[F(s)=t ^{n}\frac{ -e ^{-st} }{ s }-\int\limits_{}^{}\frac{ -e ^{-st} }{ s }nt ^{n-1}dt\]
you would have to do by parts again right
nope, we're done
you just keep going on forever
\[\begin{align}\mathcal{L}\{t^n\} ~&=~t ^{n}\frac{ -e ^{-st} }{ s }\Bigg|_{0}^{\infty}-\int\limits_{0}^{\infty}\frac{ -e ^{-st} }{ s }nt ^{n-1}dt\\~\\ ~&=~t ^{n}\frac{ -e ^{-st} }{ s }\Bigg|_{0}^{\infty}+\dfrac{n}{s}\int\limits_{0}^{\infty} t^{n-1}e ^{-st}dt\\~\\ &=0+\dfrac{n}{s}\mathcal{L}\{t^{n-1}\} \end{align}\]
so we get the recurrence relation : \[\large \mathcal{L}\{t^{n}\}~~=~~\dfrac{n}{s}\mathcal{L}\{t^{n-1}\}\]
isn't the first part (0-(-1))?
oh wait, the t out the front
makes it 0
yep, got it.
are you saying \[t ^{n}\frac{ -e ^{-st} }{ s }\Bigg|_{0}^{\infty} = 0-(-1)\] ?
my mistake
so how is this a recurrence relationship? Its been a while since ive done this
so the integral is actually the laplace of t^(n-1) right?
so we get the recurrence relation : \[\begin{align}\mathcal{L}\{t^{n}\}~~&=~~\dfrac{n}{s}\mathcal{L}\{t^{n-1}\}\\~\\ &=~~\dfrac{n}{s}*\dfrac{n-1}{s}\mathcal{L}\{t^{n-2}\}\\~\\ &=~~\dfrac{n}{s}*\dfrac{n-1}{s}*\dfrac{n-2}{s}\mathcal{L}\{t^{n-3}\}\\~\\ &=\cdots\\~\\ &=~~\dfrac{n}{s}*\dfrac{n-1}{s}*\dfrac{n-2}{s}\cdots*\dfrac{2}{s}*\dfrac{1}{s}*\mathcal{L}\{t^{0}\}\\~\\ &=~~\dfrac{n!}{s^n}*\mathcal{L}\{t^{0}\}\\~\\ \end{align}\]
could you simplify that more by saying that the laplace transform of t^0 is just the step change of the process; 1/s?
\[\frac{ n! }{ s ^{n+1} }\]
I think so, but im not so good with control theory...
yeah that would make sense because that is the final laplace transform of t^n
so im confused why you have: 2/s * 1/s *LT(t^0) why is there a 1/s factor
if psble send me the meeting id, il explain it quick
can you only run this via pc? i have a mac
oh wait then
.exe is pc only i think
what do i do?
553 005 551
is the id
message me the password
thanks for that! i'll keep working on it. cheers for the interactive help
np :) check this video when free http://ocw.mit.edu/courses/mathematics/18-03-differential-equations-spring-2010/video-lectures/lecture-19-introduction-to-the-laplace-transform/
i'm doing a course in process control, and unfortunately we never came across laplace transforms in engineering math (my degree only uses 1st semester engineering math not second) but more learning the better!
you can also take an easy one \(\mathcal L \{1\} = \int_{0}^{\infty} e^{-st} dt = \frac{1}{s}\) and diffeentiate it \(\large \frac{d}{ds} \int_{0}^{\infty} e^{-st} dt = \int_{0}^{\infty} \frac{d}{ds}e^{-st} dt = \frac{d}{ds} \frac{1}{s}\) over and over
Join our real-time social learning platform and learn together with your friends!