what Yp should i use if i have this eqn? y''-y'-6y=1+e^-2x
try \(Y_p = A + Be^{-2x}\)
why not Ax+B ?
isit because 1 on the RHS doesnt have an x counter part?
yes but with that guess, the exponent is disappearing on left hand side
@experimentX
i used the eqn you give but i cant find the coefficients.. i used Ax+B and i got yp= -1/6+e^-2x
i gotta go out soon ... what method are you using?
undertermined coefficients
\(y= A + Be^{-2x}\) \(y'= - 2Be^{-2x}\) \(y''= 4Be^{-2x}\) -------------------- \(y'' - y' - 6y = -6A = 1 + e^{-2x}\) \(\implies A = \dfrac{-1}{6}\)
B just disappeared... :/
should i change the initial guess @experimentX
try individually ... first for Y_p = x Y_p = e^(2x) or what ever .... then use superposition theorem.
that you are using A is giving wrong answer.
it must be Cx+D
or try y = Ax+B + Ce^( .... x) good luck!! gotta go out for a while. probably won't be back.
http://en.wikipedia.org/wiki/Method_of_undetermined_coefficients#Typical_forms_of_the_particular_integral the superposition of first and second.
okay... wil try thnks @experimentX :)
\(\dfrac{-1}{6} - \dfrac{1}{5}e^{-2x}\)
wolfram is giving that^^
so that means Ax+B + ce^-2x is correct?
okay okay ... this is 1/6 http://www.wolframalpha.com/input/?i=y%27%27-y%27-6y%3D1
the other is http://www.wolframalpha.com/input/?i=y%27%27-y%27-6y%3De%5E%28-2x%29
hmm.. does this mean it cant be solved by undetermined coefficients? do method power series apply on this?
\(Y_p = A + Bxe^{-2x}\)
since the first guess \(Y_p = A + Be^{-2x}\) didnt work, we make a new guess \(Y_p = A + Bxe^{-2x}\)
okay ill try
yeah! i got it! yay! xD
thanks! but, i'm not really sure how to guess what's the right Yp.. is there some kind of a clue?
\(y = A + Bxe^{-2x}\) \(y' = Be^{-2x} - 2Bxe^{-2x}\) \(y'' = -2Be^{-2x} - 2Be^{-2x} + 4Bxe^{-2x}\) ----------------------------------------------- \(y'' - y' - 6y = 6Bxe^{-2x} - 5Be^{-2x} - 6A - 6Bxe^{-2x} = 1 + e^{-2x}\)
\(y'' - y' - 6y = - 5Be^{-2x} - 6A = 1 + e^{-2x}\) \(A = \dfrac{-1}{6}\) \(B = \dfrac{-1}{5}\)
yup! i got -1/6 and -1/5 too! :D
it is the right solution as plugging it in satisfies the right hand side, right ?
oh you're asking how to make initial guess for \(Y_p\) ?
yeah.. i'm not really sure whenever i guess..
http://tutorial.math.lamar.edu/Classes/DE/UndeterminedCoefficients.aspx#Second_UnderCoeff_Ex8c
my prof have this solution but im not really sure how he does it
how does he know that it precicely the solution for the complementary equation?
you fine wid solving the complementary solution, right ?
complementary means the chaaracteristic equation?
yup solution from characteristic equaiton
\(y'' + 2y' + y = xe^{-x}\) complementary solution : \(y_c(x) = c_1e^{-x} + c_2xe^{-x}\)
complementary solution the solution for \(y'' + 2y' + y = 0\)
yeah because it has 2 distinct roots
ohh equal
\(y'' + 2y' + y = xe^{-x}\) \(y_c(x) = c_1e^{-x} + c_2xe^{-x}\) To find the particular solution, look at right hand side and make a guess
\(y'' + 2y' + y =\color{red}{ xe^{-x}}\)
the guess for polynomial \(\color{red}{x}\): \(Ax + B\) the guess for exponent \(\color{red}{e^{-x}}\) : \(Ce^{-x}\) so the overall guess is : \((Ax+B)(Ce^{-x})\)
but C is just an unnecessary constant, so so the overall guess is : \((Ax+B)e^{-x}\)
first guess : \(Y_p = (Ax+B)e^{-x}\)
fine till here ?
yup! :D thats my guess too.. but why it isnt right?
it is right, and it is indeed a solution. but the bad thing is it is `part of ur complementary solution family`. it is not constrained to the `right hand side specific function`.
\(y_c(x) = c_1e^{-x} + c_2xe^{-x}\) plugin \(c_1 = B, ~ c_2 = A\) above^
you get back ur first guess, right ?
so ur `first guess` is useless since its already part of complementary solution. u need to make a guess thats not part of `complementary solution family`
ohh.. i see, so thats why he kept adding x then? i noticed, is it right if i say, Axe^-x is identical to xe^-x, and (Ax^2+Bx)e^-x is identical (the B part which is Bxe^-x) that's why its useless?
yesss lets see why second guess is useless
first guess : \(Y_p = (Ax+B)e^{-x} \) second guess : \(Y_p = (Ax^2+ Bx)e^{-x} \)
how is the second guess part of complementary solution itself ?
the Bxe^-x ? haha
yes, complementary solution : \(y_c(x) = c_1e^{-x} + c_2xe^{-x}\) plugin \(c_1 =A= 0, ~~ c_2 = B\) u get back ur second guess
so the second guess is bad too
anyway going back to our previous question.. y"-y'-6y=1+e^-3x you put A because RHS is 1 (a constant) and you put Bxe^-3x so that it wont be the same as in the complementary solution?
for the original question, complementary solution : \(y_c(x) = c_1e^{-2x} + c_2e^{3x}\)
diff eq : \(y''-y'-6y=1+e^{-2x}\) our first guess : \(y_p = A + Be^{-2x}\)
clearly, when this is part of complementary solution family wid \(c_1 = B, ~c_2 = A = 0\)
its a bad guess
i see now.. okay, how about for this: involving trig: 9y''+y=e^-3xsin(x/3)+ xcos(x/3) my guess would be, Ae^-3xsin(x/3)+ Be^-3xcos(x/3) what do you think?
\[9y"+y=e ^{-3x}\sin (\frac{ x }{ 3 }) + xcos(\frac{ x }{ 3 })\]
complementary solution : \(y_c = c_2 \sin (\frac{x}{3}) + c_1\cos(\frac{x}{3}) \)
oh it would be the same then, so means i need to add x somewhere
guess for sin(kt) : Asin(kt) + Bcos(kt)
I think u wil get 8 variables to solve
woow why is that?
\(9y"+y=e ^{-3x}\sin (\frac{ x }{ 3 }) + x\cos(\frac{ x }{ 3 }) \) \(y_p=[Ae ^{-3x}] [B\sin (\frac{ x }{ 3 }) + C\cos(\frac{ x }{ 3 }) ] + [Dx+E][F\sin (\frac{ x }{ 3 }) + G\cos(\frac{ x }{ 3 })] \)
doesnt look pleasant to solve so many constants ;/
haha yeah xD no need to solve these just looking for Yp haha xD
btw im not sure how u got your guess honestly
i used my table... ``` function : guess x : Ax + B e^(kx) : Ae^(kx) sin(kx) : Asin(kx) + Bcos(kx) cos(kx) : Asin(kx) + Bcos(kx) ```
ooohh!! this useful! haha thank you! :D
np :)
Join our real-time social learning platform and learn together with your friends!