Eigenvalue Method for Systems of ODEs

Core Idea

To solve y' = Ay, find eigenvalues λ and eigenvectors v of A. Each eigenvalue-eigenvector pair gives a solution y = e^λtv. For complex eigenvalues, extract real and imaginary parts to form real-valued oscillating solutions.

How It's Best Learned

Work through 2×2 systems step-by-step: compute det(A - λI) = 0, find λ, solve (A - λI)v = 0 for v. Construct the general solution and verify by substitution.

Common Misconceptions

• Forgetting that real eigenvector entries are required for real-valued solutions; complex eigenvalues give oscillations. - Not recognizing that repeated eigenvalues may not have enough linearly independent eigenvectors (generalized eigenvectors needed). - Confusing eigenvectors of A with solutions to the ODE system.

Explainer

To solve a single first-order linear ODE y' = ay, you know the answer is y = Ce^at — an exponential, where a is the coefficient. The eigenvalue method generalizes this to a system of n coupled equations written as y' = Ay, where y is a vector of n unknown functions and A is an n×n matrix. The key insight is the same: look for solutions of the form y = e^λtv, where λ is a scalar and v is a constant vector. Substituting y = e^λtv into y' = Ay gives λe^λtv = Ae^λtv, and dividing by the nonzero scalar e^λt gives Av = λv. This is exactly the eigenvalue equation from linear algebra: v must be an eigenvector of A with eigenvalue λ.

So the method is: find the eigenvalues and eigenvectors of the matrix A, then construct solutions. For a 2×2 system, you compute det(A − λI) = 0 to find two eigenvalues λ₁ and λ₂. For each eigenvalue λᵢ, you solve (A − λᵢI)v = 0 to find the corresponding eigenvector vᵢ. Each pair gives an independent solution y = e^λᵢtvᵢ. The general solution is their linear combination: y = c₁e^λ₁tv₁ + c₂e^λ₂tv₂. You determine the constants c₁ and c₂ from initial conditions.

When the eigenvalues are complex — which happens when the characteristic polynomial has no real roots — the solutions still work, but you must extract real-valued solutions. If λ = α + βi is a complex eigenvalue with eigenvector v = p + qi (where p and q are real vectors), then the complex solution e^λtv expands using Euler's formula: e^(α+βi)t(p + qi) = e^αt[(cos βt)p − (sin βt)q] + ie^αt[(sin βt)p + (cos βt)q]. The real and imaginary parts are each real-valued solutions, and together they replace the pair of complex solutions. This is why complex eigenvalues produce oscillatory behavior in the system — the sin and cos terms encode rotations in the solution space.

The eigenvalue method works because the exponential structure e^λtv perfectly matches the structure of the system y' = Ay — the derivative of an exponential is proportional to itself, and a matrix-vector product with an eigenvector is proportional to the same vector. When A has n linearly independent eigenvectors (which is guaranteed if all eigenvalues are distinct, and is the usual case), the n independent solutions span the full solution space, and every solution is a linear combination of them. This is the diagonalization of the system: in the eigenvector basis, A acts by simply scaling each component, so the coupled system decouples into n independent scalar equations, each solved by an exponential.