How to find the particular solution from the differential equation?
(1-t)y''+ty'-y=2(t-1)^2e^-t y_1 and y_2 are given as e^t and t respectively
By using determining coefficient. \[y_{p} = e ^{-t}(At^2+Bt+D)\] \[y'_{p} = e ^{-t}(2At+B-At^2-Bt-D)\] \[y''_{p} = e ^{-t}(2A-4At-2B+At^2+Bt+D)\] I substituted these back into the original equation, but I am not sure of my answer.
http://www.sosmath.com/diffeq/second/variation/variation.html isn't this the method you want to use?
i didn't attempt the variation of parameters.
well if you wanted to go with the method suggested... \[y''+\frac{t}{1-t}y'-\frac{1}{1-t} y=\frac{2(t-1)^2e^{-t}}{1-t} \\ y''+\frac{t}{1-t}y'-\frac{1}{1-t}y=-\frac{2(t-1)^2e^{-t}}{t-1} \\ y''+\frac{t}{1-t}y'-\frac{1}{1-t}y=-2(t-1)e^{-t} \\ y_p=u_1 y_1+u_2 y_2 \\ y_p=u_1 e^t+u_2 t \\ \text{ we want to solve the following system } \\ u_1' e^{t}+u_2't=0 \\ u_1'e^{t}+u_2'=-2(t-1)e^{-t}\] just trying to follow their method there... \[\text{ first equation solved for } u_2' \text{ gives } \\ u_2'=-\frac{u_1' e^{t}}{t} \\ \text{ plug this into second equation } \\ u_1'e^{t}-\frac{u_1'e^{t}}{t}=-2(t-1)e^{-t} \\ \text{ multiply both sides by } t \\ t e^{t} u_1'-e^{t} u_1'=-2t (t-1)e^{-t} \\ \text{ solve for } u_1' \\ u_1'=\frac{-2t(t-1)e^{-t}}{te^{t}-e^{t}} \\ u_1' =\frac{-2t (t-1)e^{-t}}{e^t(t-1)} \\ u_1'=\frac{-2te^{-t}}{e^{t}} \\ u_1'=-2te^{-2t} \\ u_1=\frac{1}{2}e^{-2t}(2t+1)+C \\ \text{ then you can plug into } \\ u_2'=-\frac{u_1'e^{t}}{t} \text{ and find } u_2\]
u2 is a bit easier to find (no integration by parts required for that one)
Thanks. The first method I was using was time consuming.
I'm checking my work
it looks good by that method...
neat method
did you find u_2 yet?
\[u_2'=-\frac{u_1'e^{t}}{t} \\ u'_1=-2te^{-2t} \text{ plug this in} \\ u_2'=-\frac{-2te^{-2t} e^{t}}{t}\]
yes, u_2 is 2e^-t+another constant.
-2e^(-t)+c right?
i meant +ve
The wolf gave me something else. I will look at this later. I'm exhausted.
so the particular solution is: \[y_p=u_1 e^t+u_2 t \\ y_p=[\frac{1}{2}e^{-2t}(2t+1)+k_1]e^{t}+[-2e^{-t}+k_2]t\]
and we should be able to make our solution look like what they have
we just have to do some distributing and collecting like terms
\[y_p=\frac{1}{2}e^{-t}(2t+1)+k_1e^{t}-2te^{-t}+k_2 t \\ y_p=te^{-t}+\frac{1}{2}e^{-t}+k_1e^{t}-2te^{-t}+k_2t \\ \\ y_p=-te^{-t}+\frac{1}{2}e^{-t}+k_1e^{t}+k_2t\]
the only like terms was t*exp(-t) and -2t*exp(-t)
this is exactly what wolfram has
@freckles Thanks so much. I want to look this over.
k ask any questions
but just so you I followed that method exactly on that one page I refer to you up there
Join our real-time social learning platform and learn together with your friends!