6.5-exam3-2005-strang¶
Quiz 3, 2005 (strang) · Section 6.5: Solving Linear Differential Equations Source: exam3_1.pdf, pp. 96, 98
Problem¶
3 (22 pts.) Suppose the 3 by 3 matrix \(A\) has independent eigenvectors in \(Ax_1 = \lambda_1x_1\), \(Ax_2 = \lambda_2x_2\), \(Ax_3 = \lambda_3x_3\). (Those \(\lambda\)'s might not be different.)
(a) (11 points) Describe the general form of every solution \(u(t)\) to the differential equation \(\frac{du}{dt} = Au\). (The answer \(e^{At}u(0)\) does not use the \(\lambda\)'s and \(x\)'s.)
(b) (11 points) Starting from any vector \(u_0\) in R\(^3\), suppose \(u_{k+1} = Au_k\). What are the conditions on the \(x\)'s and \(\lambda\)'s to guarantee that \(u_k \to 0\) (as \(k \to \infty\))? Why?
Solution
(a) The general solution to the differential equation \(\frac{du}{dt} = Au\) is \[ u(t) = c_1 e^{\lambda_1 t} x_1 + c_2 e^{\lambda_2 t} x_2 + c_3 e^{\lambda_3 t} x_3, \] where \(c_1, c_2, c_3\) are arbitrary constants. (b) Since the vectors \(x_1, x_2, x_3\) are independent, they form a basis for \(R^3\). It follows that we may write any vector \(u_0 \in R^3\) as a linear combination of these vectors: \(u_0 = a_1 x_1 + a_2 x_2 + a_3 x_3\). Repeatedly applying the matrix \(A\) we obtain \[ u_k = A^k u_0 = \lambda_1^k a_1 x_1 + \lambda_2^k a_2 x_2 + \lambda_3^k a_3 x_3. \] If we want the limit as \(k\) goes to infinity of the vectors \(u_k\) to be zero, then all the limits \(\lim \lambda_i^k\) must be zero. It follows that we necessarily have \(-1 < \lambda_i < 1\) for all \(i\)'s.Keywords: eigenvalues, eigenvectors, differential equation, stability, matrix powers