时域中的线性近似

State space representation effectively captures nonlinear system dynamics by using state variables to describe position and velocity. This approach is particularly useful for systems where variables and parameters vary with time or operating conditions, such as pendulums or springs. State space form enables systematic linearization around equilibrium points and facilitates analysis using linear control techniques.

Taylor series approximation linearizes nonlinear equations by expanding them around an equilibrium point using small perturbations of state variables. The nonlinear terms are replaced by their linear counterparts, yielding linear state equations. This technique allows engineers to apply linear system theory and control strategies to inherently nonlinear systems like pendulums and spring systems.

Equilibrium points define the operating conditions around which linearization occurs. By perturbing state variables about their equilibrium values, nonlinear dynamics are approximated linearly. For spring systems, the equilibrium force at position x0 is calculated to formulate the linearized differential equation, enabling accurate local system behavior analysis.

State variables for a simple pendulum are chosen to represent the system's position and velocity. With the center of mass at half the pendulum's length, state variables capture angular position and angular velocity. These variables lead to state equations describing the system's time evolution under gravitational force and applied torque effects.

Converting linearized equations into vector-matrix form provides a comprehensive, standardized representation suitable for analysis using linear control and estimation techniques. This format enables systematic design and implementation of control strategies. Vector-matrix representation facilitates computational methods and allows engineers to apply established linear system theory to complex mechanical systems.

Both systems use Taylor series expansion around equilibrium, but spring systems focus on translational motion with nonlinear spring force, while pendulums involve rotational dynamics with gravitational torque. For springs, equilibrium force at x0 is calculated; for pendulums, torque balance is summed. Both yield linear state equations valid for small perturbations around equilibrium positions.

Small perturbations ensure that Taylor series truncation at linear terms remains accurate. Large deviations from equilibrium introduce significant higher-order nonlinear terms that the linear approximation cannot capture. Restricting analysis to small perturbations validates the linearized model's predictions and ensures control strategies designed using linear theory remain effective.