1
$\begingroup$

Given this system

$$\begin{align} \ddot{y_1} & = -y_1 + y_2 + 1 - a \\ \ddot{y_2} & = y_1 - 2 y_2 + y_3 + 1 \\ \ddot{y_3} & = y_2 - y_3 + 1 + a \\ \end{align} $$

putting this into matrix form gives $\ddot{\textbf{y}} = A \textbf{y} + \textbf{g}$ wherein $\textbf{y}, \textbf{g}$ are the appropriate $3\times1$ matrices for the system and $A$ is the corresponding coefficient matrix. One problem that I am having is that when the eigenvalues are computed ($\lambda_1 = 0 \, (\text{multiplicity of 2}), \lambda_2 = \pm i, \lambda_3 = \pm i \sqrt{3}$), one of the corresponding eigenvectors for $\lambda_1$ doesn't exist. As an illustration:

$$\left(\begin{matrix} -1 & 1 & 0 \\ 1 & -2 & 1 \\ 0 & 1 & -1 \\ \end{matrix} \right) \left(\begin{matrix} v_1 \\ v_2 \\ v_3 \\ \end{matrix} \right) = \left(\begin{matrix} 0 \\ 0 \\ 0 \end{matrix} \right)$$

gives $\textbf{v} = \left(1 \quad 1 \quad 1 \right)^T$ as the first eigenvector. Solving for the other eigenvector:

$$\left(\begin{matrix} -1 & 1 & 0 \\ 1 & -2 & 1 \\ 0 & 1 & -1 \\ \end{matrix} \right) \left(\begin{matrix} p_1 \\ p_2 \\ p_3 \\ \end{matrix} \right) = \left(\begin{matrix} 1 \\ 1 \\ 1 \end{matrix} \right)$$

Which leads to this

$$\left(\begin{matrix} -1 & 1 & 0 \\ 0 & -1 & 1 \\ 0 & 0 & 0 \\ \end{matrix} \right) \left(\begin{matrix} p_1 \\ p_2 \\ p_3 \\ \end{matrix} \right) = \left(\begin{matrix} 1 \\ 2 \\ 3 \end{matrix} \right)$$

which does not make sense. Does this mean that the second eigenvector for $\lambda_1$ does not exist? Also as an aside, after solving the homogeneous case, how do I solve for the inhomogeneous case?

$\endgroup$
  • 2
    $\begingroup$ Shouldn't you write those as first order equations first and get a $6~x~6$ from $y' = A y + g$? $\endgroup$ – Moo Oct 25 '18 at 5:02
  • $\begingroup$ @Moo Yes, however I've read from here that the characteristic equation for the $6 \times 6$ system is just $(-1)^n$ times the characteristic equation for the $3 \times 3$ system. However I do think that I would still have to use the $6 \times 6$ system for the particular solution. $\endgroup$ – Aldon Oct 25 '18 at 5:22
  • $\begingroup$ You're certainly not going to get 6 eigenvalues and eigenvectors from a $3\times3$ matrix. $\endgroup$ – Gerry Myerson Oct 25 '18 at 6:45
1
$\begingroup$

Given $$\begin{align} \ddot{y_1} & = -y_1 + y_2 + 1 - a \\ \ddot{y_2} & = y_1 - 2 y_2 + y_3 + 1 \\ \ddot{y_3} & = y_2 - y_3 + 1 + a \\ \end{align} $$

Writing this as first order equations yields

$$X' = \left(\begin{matrix} x_1' \\ x_2' \\ x_3' \\ x_4' \\x_5'\\x_6' \end{matrix} \right) = AX + g = \left(\begin{matrix} 0 & 1 & 0 & 0 & 0 & 0 \\ -1 & 0 & 1 &0&0&0\\ 0 & 0 & 0 & 1 & 0 & 0 \\ 1 & 0 & -2 & 0 & 1 & 0 \\ 0 & 0 & 0 & 1 & 0 & 1 \\ 0 & 0 & 1 & 0 & -1 & 0 \end{matrix} \right) \left(\begin{matrix} x_1 \\ x_2 \\ x_3 \\ x_4 \\x_5\\x_6 \end{matrix} \right)+ \left(\begin{matrix} 0 \\ 1-a \\ 0 \\ 1 \\ 0\\1+a \end{matrix} \right)$$

The eigenvalues of $A$ are

$$\lambda_{1, 2} = \pm~i~\sqrt{3}, ~~\lambda_{3, 4} = \pm~ i, ~~\lambda_{5,6} = 0$$

We have a deficient matrix and need to find a generalized eigenvector for $\lambda = 0$. Using Wiki Example $1$ as a possible approach, we have

$$[A - \lambda I]v_6 = [A]v_6 = v_5$$

This leads to the augmented matrix $$ RREF \left(\begin{array}{rrrrrr|r} 0 & 1 & 0 & 0 & 0 & 0 & 1 \\ -1 & 0 & 1 & 0 & 0 & 0 & 0 \\ 0 & 0 & 0 & 1 & 0 & 0 & 1 \\ 1 & 0 & -2 & 0 & 1 & 0 & 0 \\ 0 & 0 & 0 & 0 & 0 & 1 & 1 \\ 0 & 0 & 1 & 0 & -1 & 0 & 0 \\ \end{array}\right) = \left( \begin{array}{rrrrrr|r} 1 & 0 & 0 & 0 & -1 & 0 & 0 \\ 0 & 1 & 0 & 0 & 0 & 0 & 1 \\ 0 & 0 & 1 & 0 & -1 & 0 & 0 \\ 0 & 0 & 0 & 1 & 0 & 0 & 1 \\ 0 & 0 & 0 & 0 & 0 & 1 & 1 \\ 0 & 0 & 0 & 0 & 0 & 0 & 0 \\ \end{array} \right) $$

From this, we have the generalized eigenvector

$$v_6 = \left(\begin{matrix} 0 \\ 1 \\ 0 \\ 1 \\0\\1 \end{matrix} \right) $$

For the other four eigenvalues, we can use the typical methods to find linearly independent eigenvectors.

Next, to solve the nonhomogeneous system, refer to the various methods here.

$\endgroup$
1
$\begingroup$

$$ A=\pmatrix{-1&1&0\\1&-2&1\\0&1&-1} $$ has characteristic polynomial $$-(q+1)^2(q+2)+2(q+1)=(q+1)(2-2-3q-q^2)=-q(q+1)(q+3)$$ which gives an an eigen-decomposition $UDU^T$ with $$ D=\pmatrix{-3&0&0\\0&-1&0\\0&0&0}\text{ and } U=\pmatrix{ \frac1{\sqrt6}& \frac1{\sqrt2}&\frac1{\sqrt3}\\ -\frac2{\sqrt6}& 0 &\frac1{\sqrt3}\\ \frac1{\sqrt6}&-\frac1{\sqrt2}&\frac1{\sqrt3} } $$ With the coordinate change to the basis in the columns of $U$, the original order 2 system decouples into 2 oscillation equations $\ddot x_1+3x_1=c_1$, $\ddot x_2+x_2=c_2$ and a simple one of the form $\ddot x_3=c_3$.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.