Model Predictive control for beginners by Anthony Rossiter

Chapter 1: Classical control and weaknesses

It is not the purpose of this book to teach classical control or to give a comprehensive view of the control techniques used in industry. Nevertheless, it is useful to begin from a brief summary of the sorts of techniques which are commonly used. For convenience and clarity, these will be presented in the simplest forms. This is followed by brief illustrations of the types of scenarios where such controller structures are difficult to tune effectively and robustly.

  1. Common controller structures.
  2. Non-minimum phase zeros.
  3. Incorporating constraints.
  4. Systems with delays.
  5. Multivariable systems.

1.1: Common controller structures

Gives a very brief introduction to common control structures such as PID and lead/lag.

Return to top of this page

1.2: Non-minimum phase zeros

Gives an introduction to the efficacy of classical control structures when used with processes having right half plane zeros. Demonstrates that, at times, simple classical structures are too limited in flexibility to obtain the desired performance.

Return to top of this page

1.3: Systems with delays

Many real processes include a significant delay. These two files introduce delays and the impact on closed-loop performance and also a common industrial solution, the so called smith predictor.

Return to top of this page

1.4: Constraints

Discusses the impact of system constraints on closed-loop behaviour and demonstrates that a failure to include these systematically can have catastrophic effects. It is noted that with classical control constraints are an afterthought rather than included systematically.


Return to top of this page

1.5: Multivariable processes

Multivariable processes are more difficult to control due to the interaction between each input and output. This short file illustrates the issue concisely and exposes the difficulties with using classicial control approaches, but without considering in detail the potential solutions.


Return to top of this page