system of linear ODEs

Real systems rarely have a single moving part. Two masses coupled by springs, two interacting populations, the currents in a multi-loop circuit, the concentrations in a chain of reacting chemicals — each is governed not by one equation but by several coupled equations that must be solved together. A system of linear ODEs is the framework for all of these.

Stacking the unknowns into a vector x = (x_1, ..., x_n) and the coefficients into a matrix A, a linear system writes compactly as x' = A x (homogeneous) or x' = A x + g(t) (forced). This vector form is more than notation: it reveals that the whole theory of single equations carries over with matrices in place of numbers. Solutions superpose; the general homogeneous solution is a combination of n independent vector solutions; and there is a structural reason for that — any higher-order linear ODE can itself be rewritten as a first-order system by treating each derivative as a new variable, so systems are the unifying form.

Systems are the language of modern dynamics and control. Writing a model as x' = A x exposes its structure to linear algebra: the eigenvalues of A determine every qualitative behavior (growth, decay, oscillation, stability), the eigenvectors give the natural modes, and the matrix exponential e^(At) gives the exact solution. This is also the form every numerical ODE solver expects, so reducing any problem to a first-order system is the standard first step in computation.