(I) Fundamental Matrix Solution
Consider the homogeneous linear system \(\vec{x}' = \mathbf{A}\vec{x}\). Let \(\vec{x}_1(t), \dots, \vec{x}_n(t)\) be \(n\) linearly independent solutions. A **fundamental matrix** \(\Phi(t)\) is constructed by placing these solutions as its columns:
Relationship Between Fundamental Matrices
A fundamental matrix is not unique. If \(\Psi(t)\) is any other fundamental matrix for the same system, then each column of \(\Psi(t)\) is a linear combination of the columns of \(\Phi(t)\). Therefore, there exists a constant \(n \times n\) non-singular matrix \(\mathbf{M}\) such that:
Theorem 1: Fundamental Matrix Solutions
The unique solution to the Initial Value Problem \(\vec{x}' = \mathbf{A}\vec{x}, \vec{x}(0) = \vec{x}_0\) is given by:
Solve: \( x' = 4x + 2y, y' = 3x - y \) with \( x(0) = 1, y(0) = -1 \).
Step 1: Fundamental Matrix. Two solutions are \(\vec{x}_1 = \begin{bmatrix} e^{-2t} \\ -3e^{-2t} \end{bmatrix}, \vec{x}_2 = \begin{bmatrix} 2e^{5t} \\ e^{5t} \end{bmatrix}\). Thus:
Step 2: Solution. Using \(\vec{x}(t) = \Phi(t)\Phi(0)^{-1}\vec{x}_0\):
(II) Exponential Matrices
Matrix exponentials are defined by the series:
Let \(\mathbf{A} = \begin{bmatrix} 2 & 3 & 4 \\ 0 & 2 & 6 \\ 0 & 0 & 2 \end{bmatrix}\). Decompose into \(\mathbf{A} = 2\mathbf{I} + \mathbf{B}\) where \(\mathbf{B}\) is nilpotent.
Step 1: Identity Part. The exponential of the identity multiple is:
Step 2: Nilpotent Part. Since \(\mathbf{B}^2 = \begin{bmatrix} 0 & 0 & 18 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \end{bmatrix}\) and \(\mathbf{B}^3 = \mathbf{0}\):
Step 3: Final Matrix. Since \(2\mathbf{I}t\) and \(\mathbf{B}t\) commute:
(III) Matrix Exponential Solutions
Theorem 2: Exponential Solutions
The unique solution to \(\vec{x}' = \mathbf{A}\vec{x}, \vec{x}(0) = \vec{x}_0\) is:
This implies that
For \(\mathbf{A} = \begin{bmatrix} 4 & 2 \\ 3 & -1 \end{bmatrix}\), using \(\Phi(t)\) from Example 1:
Solve \(\vec{x}' = \begin{bmatrix} 2 & 5 \\ 0 & 2 \end{bmatrix}\vec{x}, \vec{x}(0) = \begin{bmatrix} 4 \\ 7 \end{bmatrix}\).
Decompose \(\mathbf{A} = 2\mathbf{I} + \mathbf{B}\). Then \(e^{\mathbf{A}t} = e^{2\mathbf{I}t}e^{\mathbf{B}t}\):
Final Particular Solution: