Lecture slides: control-system-design-objectives.pdf

Reading suggestion: KJA_pp206_235.pdf

Feedback, if properly designed, can be used to make complex systems behave predictably, even when operating in noisy and uncertain environments. We start by introducing and discussing the common objectives of feedback design.

Systems in the real world operate in unpredictable environments. Designing to mitigate the effect of such disturbances is an important phase of control system design.

Feedback is used to reduce uncertainty, but the process of feedback also introduces additional uncertainty through measurement noise. We discuss this, and some common approaches to reducing its effect.

System models, when they are available at all, are rarely perfect, often only approximately describing the true input-output behaviour of a system. Yet despite this, simple control systems (typically PID) are able to obtain robust and predictable performance. We address this apparent contradiction by studying the robustness properties of feedback loops.