From differential equations to matrix form
State-space representation captures a system's dynamics as a vector of state variables and a set of first-order matrix equations. Instead of a single high-order differential equation, the system is described as dx/dt = Ax + Bu, where x is the state vector, u is the input, and A and B are matrices encoding the relationships. The output y is computed from y = Cx + Du, where C and D are output matrices. This matrix form is natural for computers and control design tools, and it extends easily to multiple inputs and outputs (MIMO systems).
Advantages for analysis and design
State-space form makes it easy to assess controllability, observability, and stability by examining matrix ranks and eigenvalues. Pole placement (choosing where system eigenvalues lie) is a direct design method using state feedback. The matrices also encode information that transfer functions hide, such as the order of the system and the presence of unstable modes that might cancel in a simplified transfer function. Most modern control design software works in state-space form and converts the results back to transfer functions or PID parameters for implementation.