de-CH
utf-8
math math-format
IVP and Convergence
ev-ew-01-01
multiple
1048576
randRange(1,12) randRange(-12,-1) randRange(1,12) randRange(1,12) randRange(1,12) randRange(1,D*Y/X-1) L - (X*B)/Y X/Y *(L - D) fractionReduce(Y*L - X * B,Y) fractionReduce(X *(L - D),Y) fractionReduce(D*Y*L - B*X *L,Y) AN*D - CN * B fractionReduce(Y*D + Y*L - X * B,Y) fractionReduce(Y*D - X * B,Y)

The matrix A = \begin{pmatrix} A & B\\ C & D \end{pmatrix} defines a system y' = Ay.

Compute the entry {\color{red}B}, such that the solution of the system starts with y(0) = \begin{pmatrix} Y*Z \\ {\color{red}B} \end{pmatrix} and that \displaystyle \lim_{t \to \infty} y(t) = \begin{pmatrix} 0 \\ 0 \end{pmatrix} holds.

EV \color{red} B = X*Z

At first, compute eigenvalues {\color{blue}\lambda_{1,2}} as zeroes:

\lambda^2 - T \cdot \lambda + S = \left(\lambda - {\color{blue}negParens(L)} \right) \cdot \left(\lambda - {\color{blue}L2}\right) = \left(\lambda - {\color{blue}\lambda_1} \right) \cdot \left(\lambda - {\color{blue}\lambda_2} \right) .

With corresponding eigenvectors v_1 und v_2, we get the general solution

y: \mathbb R \to \mathbb R^2 mit y(t) = C_1 \cdot e^{ {\color{blue}L} \cdot t} \cdot v_1 + {\color{orange}C_2}\cdot e^{{\color{blue}\lambda_2} \cdot t} \cdot v_2.

Thus we must find {\color{red}B} such that y(0) = \begin{pmatrix} Y*Z \\ {\color{red}B} \end{pmatrix} and in addition \color{orange}C_2 = 0, as {\color{blue}\lambda_2} = L2 \geq 0.

Dies garantiert die Konvergenz \displaystyle \lim_{t \to \infty} y(t) = \lim_{t \to \infty} C_1 e^{ {\color{blue}L} \cdot t} \cdot v_1 = \begin{pmatrix} 0 \\ 0 \end{pmatrix}.

With the given numbers this yields to:

y(0) = \begin{pmatrix} Y*Z \\ {\color{red}B} \end{pmatrix} = C_1 \cdot e^{ {\color{blue}L} \cdot 0} \cdot v_1 = C_1 \cdot v_1.

To proceed further, we need to determine the eigenvectors:

From A v_1 = {\color{blue}\lambda_1} v_1 we conclude that the eigenvectors are of the form \alpha\cdot \begin{pmatrix} Y \\ X \end{pmatrix} with 0 \neq \alpha \in \mathbb R.

The 1st coordinate give Y*Z = C_1 \cdot Y and hence Z = C_1.

This Z = C_1 in the 2nd coordinate equation leads to {\color{red}B} = Z*X.