Feedback Control

Open-Loop Control vs. Feedback Control

Recall that the basic objectives of control are to:

• Track a given reference
• Reject disturbances
• Meet performance specs

Open-loop control:

• Cheaper/easier to implement (no sensor) required
• Does not destabilize the system. This means that if both $K$ and $P$ are stable (all poles are in the open left-half plane), $$\frac{Y}{R}=KP$$ is also stable, and we have $$\{\text{poles of }KP\}=\{\text{poles of }K\}\cup \{\text{poles of }P\}$$

Feedback control:

• More difficult/expensive to implement (requires a sensor and an information path from controller to actuator)
• May destabilize the system as $$\frac{Y}{R}=\frac{KP}{1+KP}$$ has new poles, which may be unstable
• However, feedback control is the only way to stabilize an unstable plant.
• Reduces steady-state error to disturbances
• Reduces steady-state sensitivity to model uncertainty (parameter variations)
• Improves time response