ODE: Eigenvalue Method for Systems

Mastering the eigenvalue method for systems of ordinary differential equations (ODEs) is a cornerstone skill in engineering. It transforms the daunting task of solving multiple, interdependent equations into a structured, algebraic procedure. This method is indispensable for analyzing dynamic systems, from mechanical vibrations and electrical circuits to control theory and structural dynamics, providing not just solutions but deep insight into system behavior.

Understanding the Core Form and Eigenvalue Fundamentals

We begin with a homogeneous linear system of first-order ODEs with constant coefficients, which can be written compactly in matrix form: $${\mathbf{x}}^{\prime }=A\mathbf{x}$$ Here, $\mathbf{x}$ is a vector of unknown functions (e.g., $[x_1(t),x_2(t){]}^T$) and $A$ is an $n\times n$ constant matrix. The eigenvalue method proposes that solutions take the form $\mathbf{x}=\mathbf{v}{e}^{\lambda t}$, where $\mathbf{v}$ is a constant vector and $\lambda$ is a constant scalar. Substituting this guess into the system leads to the fundamental eigenvalue problem: $$A\mathbf{v}=\lambda \mathbf{v}$$ This equation has nontrivial solutions for $\mathbf{v}$ only if $\det (A-\lambda I)=0$. The roots of this characteristic equation are the eigenvalues $\lambda$, and their corresponding vectors $\mathbf{v}$ are the eigenvectors. The nature of these eigenvalues—real distinct, complex, or repeated—directly dictates the form of the general solution to the system.

Case 1: Real and Distinct Eigenvalues

This is the most straightforward case. If an $n\times n$ matrix $A$ yields $n$ real distinct eigenvalues ${\lambda }_1,{\lambda }_2,...,{\lambda }_n$ with corresponding eigenvectors ${\mathbf{v}}_1,{\mathbf{v}}_2,...,{\mathbf{v}}_n$, the general solution is a linear combination of the fundamental solutions: $$\mathbf{x}(t)=c_1{\mathbf{v}}_1{e}^{{\lambda }_{1}t}+c_2{\mathbf{v}}_2{e}^{{\lambda }_{2}t}+...+c_n{\mathbf{v}}_n{e}^{{\lambda }_{n}t}$$ Each term represents a mode of the system's behavior. For example, in a two-equation system modeling a coupled mass-spring system, a positive eigenvalue indicates an unstable, exponentially growing mode, while a negative eigenvalue indicates a decaying, stable mode. The constants $c_1,c_2,...,c_n$ are determined by the system's initial conditions.

Case 2: Complex Conjugate Eigenvalues

Complex eigenvalues always occur in conjugate pairs for real matrices. Suppose $\lambda =\alpha \pm \beta i$ (where $i=\sqrt{-1}$) are a pair of complex conjugate eigenvalues with corresponding complex eigenvectors $\mathbf{v}=\mathbf{a}\pm \mathbf{b}i$. While $c_1\mathbf{v}{e}^{(\alpha +\beta i)t}+c_2\bar{\mathbf{v}}{e}^{(\alpha -\beta i)t}$ is a valid complex-valued solution, engineers require real-valued solutions. Using Euler's formula ($e^{i\theta }=\cos \theta +i\sin \theta$), we extract two linearly independent real solutions: $${\mathbf{x}}_1(t)=e^{\alpha t}[\mathbf{a}\cos (\beta t)-\mathbf{b}\sin (\beta t)]$$ $${\mathbf{x}}_2(t)=e^{\alpha t}[\mathbf{a}\sin (\beta t)+\mathbf{b}\cos (\beta t)]$$ The general real solution is then $\mathbf{x}(t)=c_1{\mathbf{x}}_1(t)+c_2{\mathbf{x}}_2(t)$. The real part $\alpha$ determines exponential growth/decay, while the imaginary part $\beta$ gives the oscillatory frequency. This case is central to analyzing damped or undamped oscillations in engineering systems.

Case 3: Repeated Eigenvalues and Generalized Eigenvectors

A repeated eigenvalue (or eigenvalue with algebraic multiplicity $m>1$) may not have enough linearly independent eigenvectors. A matrix lacking a full set of eigenvectors is called defective. In these defective matrix cases, we must find generalized eigenvectors to construct a complete solution.

For a repeated eigenvalue $\lambda$ with geometric multiplicity (number of independent eigenvectors) less than its algebraic multiplicity, we solve the chain of equations: $$(A-\lambda I){\mathbf{v}}_1=0$$ $$(A-\lambda I){\mathbf{v}}_2={\mathbf{v}}_1$$ $$(A-\lambda I){\mathbf{v}}_3={\mathbf{v}}_2$$ and so on. Here, ${\mathbf{v}}_1$ is a genuine eigenvector, and ${\mathbf{v}}_2,{\mathbf{v}}_3,...$ are generalized eigenvectors. The corresponding fundamental solutions for a repeated eigenvalue of multiplicity 2 are: $${\mathbf{x}}_1(t)={\mathbf{v}}_1{e}^{\lambda t}$$ $${\mathbf{x}}_2(t)=({\mathbf{v}}_{1}t+{\mathbf{v}}_2)e^{\lambda t}$$ For a multiplicity of 3, a third solution of the form $({\mathbf{v}}_1\frac{t^2}{2}+{\mathbf{v}}_{2}t+{\mathbf{v}}_3)e^{\lambda t}$ appears. This method systematically builds the required number of independent solutions.

Constructing the Complete General Solution

The final, complete solution for all eigenvalue types is assembled by taking a linear combination of all linearly independent fundamental solutions found from each eigenvalue. The process is systematic:

1. Find all eigenvalues by solving $\det (A-\lambda I)=0$.
2. For each eigenvalue, find its eigenvectors and, if necessary, generalized eigenvectors.
3. Write the fundamental solution set:

• Real $\lambda$: $\mathbf{v}{e}^{\lambda t}$.
• Complex pair $\alpha \pm \beta i$: $e^{\alpha t}[\mathbf{a}\cos (\beta t)-\mathbf{b}\sin (\beta t)]$ and $e^{\alpha t}[\mathbf{a}\sin (\beta t)+\mathbf{b}\cos (\beta t)]$.
• Repeated $\lambda$: Include terms with $t$ and $t^2$ factors as needed, using generalized eigenvectors.

1. The general solution is $\mathbf{x}(t)=c_1{\mathbf{x}}_1(t)+c_2{\mathbf{x}}_2(t)+...$.

This framework is universal for constant-coefficient linear systems.

Common Pitfalls

1. Forgetting to Extract Real Solutions from Complex Eigenvalues: A common error is presenting the final answer in complex form (e.g., using $e^{(\alpha +\beta i)t}$). Always remember the final engineering solution must be real-valued. Use Euler's formula to convert the complex exponential pair into real sine and cosine functions.
2. Misapplying the Repeated Root Formula: A critical mistake is attempting to use the generalized eigenvector method when the matrix is not defective. If a repeated eigenvalue has the same number of independent eigenvectors as its multiplicity (a non-defective or complete case), you simply use the standard $\mathbf{v}{e}^{\lambda t}$ solution for each eigenvector. Always check the dimension of the eigenspace (nullspace of $(A-\lambda I)$) first.
3. Incorrect Chain Order for Generalized Eigenvectors: When solving $(A-\lambda I){\mathbf{v}}_k={\mathbf{v}}_{k-1}$, the chain must start with a true eigenvector ${\mathbf{v}}_1$. The equation $(A-\lambda I){\mathbf{v}}_1=0$ must hold. Starting with an arbitrary vector will lead to an inconsistent system. Solve sequentially and verify each step.
4. Neglecting Initial Conditions for Constants: After constructing the elegant general solution, it's easy to forget the final, practical step. The vector of constants $c_1,c_2,...$ is found by applying the initial condition $\mathbf{x}(0)={\mathbf{x}}_0$. This involves solving a system of linear algebraic equations. The solution is not complete until these constants are specified for a given problem.

Summary

• The eigenvalue method reduces solving ${\mathbf{x}}^{\prime }=A\mathbf{x}$ to the algebraic problem of finding eigenvalues and eigenvectors of the matrix $A$.
• Real distinct eigenvalues yield solutions of the form $\mathbf{v}{e}^{\lambda t}$, while complex conjugate eigenvalues $\alpha \pm \beta i$ yield oscillatory solutions involving $e^{\alpha t}\cos (\beta t)$ and $e^{\alpha t}\sin (\beta t)$.
• For repeated eigenvalues in a defective matrix, you must find generalized eigenvectors to create complete solutions, which introduce polynomial terms like $t{e}^{\lambda t}$.
• The complete general solution is always a linear combination of all fundamental solution vectors found from each eigenvalue and its associated vectors.
• Always check if a repeated eigenvalue is defective (lacking enough eigenvectors) before proceeding to find generalized eigenvectors, and remember to convert complex-valued solution forms into real-valued functions for physical interpretation.