just because i needed to refresh myself on this particular method .... variation of parameters.
Variation of Parameters: the solution to the differential equation: \[y'' + py' + qy = f(x)\] can be determined if the homogenous solution \(y_h\) is known. \[(y_h)'' + p(y_h)' + q(y_h) = 0\] Let \(y_h=c_1u(x)+c_2v(x)\) such that \(c_1\) and \(c_2\) are arbitray constants, given \(u(x)\) and \(v(x)\) are linearly independant. By varing the parameters of \(y_h\), we can determine a function \(y_p\). Let g(x) and h(x) be 2 arbitraty functions and replace the arbitrary constants with them to obtain: \[y_p=g(x)u(x)+h(x)v(x)~:=~gu+hv\] Take the first and second derivatives to insert them into the equation: \[\begin{matrix}(y_p)' &=& g'u + gu' + h'v+ hv'\\ &=& gu' + hv' + (\underbrace{g'u + h'v}_{=0}) \end{matrix}\] \[(y_p)''= g'u' +gu'' + h'v' + hv'' + (\underbrace{g'u + h'v}_{=0})'\] By setting \(g'u + h'v = 0\) we give ourselves the ability to chose g or h at random; and if one is chosen the other can be determined (substitution method of solving a system of equations). Also, setting it equal to zero frees up some calculations, otherwise we would have ended up with 8 terms in the 2nd derivative. Now we can input these into the original equation: \[\begin{pmatrix} +q~gu & +q~hv &+g'u' \\ +p~gu' & +p~hv' &+h'v' \\ +gu'' & +p~hv'' & \\ \color{red}{=0}&\color{red}{=0} \end{pmatrix}=f(x)\] In the first 2 columns, we can factor out g and h, and what remains is a product that contains the homogeneous solutions -- the solutions that equates to 0 -- so that our setup reduces to simply: \[g'u'+h'v' = f(x)\] Recall that we defined something similar to this beforehand: \[g'u + h'v = 0\] We now have 2 equations in 2 unknowns (g' and h') that can be worked out using algebraic methods: substitution, cramers rule, elimination, etc. \[g'u + h'v = 0\\ g'u' + h'v' = f(x)\] When we know g' and h', we integrate to obtain g and h which were the arbitrary functions that we were looking for. thank you herbert gross; will post his video once i find it on the google again.
Join our real-time social learning platform and learn together with your friends!