91Ó°ÊÓ

In the context of differential equations, eigenvalues are critical in determining the behavior of solutions over time. When dealing with a matrix equation like \( \frac{d \mathbf{x}}{d t} = A \mathbf{x}(t) \), eigenvalues help us understand how the system evolves. In simple terms, eigenvalues are numbers that provide insight into the "stretching" or "compressing" action applied by the matrix \( A \). They tell us about the rates at which trajectories, or solutions, are moving towards or away from an equilibrium point.

To find the eigenvalues of a matrix \( A \), you'll solve the characteristic equation \( \det(A - \lambda I) = 0 \). In our example, solving \((-1 - \lambda)(4 - \lambda) = 0\), the eigenvalues are \( \lambda_1 = -1 \) and \( \lambda_2 = 4 \). These are real, distinct, and nonzero, which means that the equilibrium point at \((0,0)\) is affected by both a negative and a positive trajectory.

In this case, the eigenvalues indicate how the accompanying solutions of the differential equation respond over time. The negative eigenvalue suggests that there is a decay or reduction over time for trajectories in one direction, while the positive eigenvalue suggests growth or expansion in the other. These behaviors are foundational to analyzing the stability and type of equilibrium points.

Stability analysis is a tool used to understand how a dynamic system behaves when it's slightly perturbed or displaced from its equilibrium. Essentially, it asks: "If the system starts close to an equilibrium point, will it stay nearby or will it drift away?"

In the case of eigenvalues, stability is assessed by observing the real parts of these values. If all eigenvalues have negative real parts, the system is stable—trajectories tend to return towards the equilibrium point after a disturbance. Conversely, if any eigenvalue has a positive real part, the system is unstable because trajectories are pushed away, not returning to the equilibrium.

In our exercise, with eigenvalues \( \lambda_1 = -1 \) (negative) and \( \lambda_2 = 4 \) (positive), the system is deemed unstable. This mix of signs means that part of the system attracts trajectories toward the equilibrium, while another part pushes them away. Consequently, the stability analysis reveals that the equilibrium point at \((0,0)\) is unstable due to one eigenvalue's positive real part.

A saddle point is a specific type of equilibrium point that exhibits inherent instability. It occurs when there is a mixture of positive and negative behaviors in a system, typically represented by eigenvalues of opposite signs.

In a two-dimensional system, when the equilibrium point is a saddle point, one can visualize it as having two distinct directions of movement. In the direction of the eigenvalue with a negative sign, the motion is towards equilibrium—imagine a horse's saddle where ridges rise up. In contrast, in the direction of the positive eigenvalue, motion is away from equilibrium—like sliding down the sides of the saddle.

Our given system with matrix \( A = \begin{bmatrix} -1 & 0 \ 0 & 4 \end{bmatrix} \) produces eigenvalues \( \lambda_1 = -1 \) and \( \lambda_2 = 4 \). This configuration leads to a saddle point at \((0,0)\) because the eigenvalues have different signs. Through this classification, students learn a crucial insight: contrasting eigenvalues indicate that certain directions are stable while others are unstable, resulting in the familiar saddle point behavior. A saddle point is inherently unstable, making this a pivotal observation in analyzing differential equations.