Control Systems Overview
A control system drives a physical process toward desired behavior by shaping its inputs based on goals and measurements.
What a Control System Does
A control system takes a plant (the physical process to be governed), a reference (what we want it to do), and a controller (logic that computes the commanded input). The controller's job is to make the plant's output track the reference despite disturbances, model error, and noise.
Every control problem carries a set of standing tensions: speed of response versus overshoot, tracking accuracy versus control effort, and performance versus robustness to modeling error. Good design is the deliberate management of these trade-offs, not the elimination of any single one.
Core vocabulary
- Plant: the system under control, often modeled by differential equations.
- Actuator: converts controller commands into physical action.
- Sensor: measures the output, always with some noise.
- Disturbance: an unwanted input that perturbs the plant.
- Setpoint or reference: the target value for the output.
Two broad architectures
In an open-loop system the controller acts on the reference alone, with no measurement of the result. In a closed-loop system the measured output is fed back and compared to the reference, so the controller reacts to error. Feedback is what gives control systems their power to reject disturbances and tolerate imperfect models.
Where it is used
Control theory underlies flight autopilots, chemical reactors, robotics, power grids, and disk drives. In fusion engineering, feedback stabilizes plasma position, shape, and current: the Hyperion breeder concept is a design-and-simulation spherical tokamak whose vertical position at negative triangularity would require active magnetic control, a classic unstable-plant feedback problem.
The rest of this section develops the mathematics that makes such design systematic: transfer functions, stability tests, state-space models, estimators, and optimal and robust controllers.