MSc Advanced Control and Systems Engineering / Course details

The unit covers the following topics:

Part A - Digital Control

• Motivation for digital control theory, including computer-based control.
• Discrete representation of continuous systems: discrete transfer functions, the z transform, and state-space modelling.
• Stability analysis.
• Control system design concerns in practice (e.g. sampling rate selection, computational time) in the discrete domain. 
• Classical anaysis in the discrete domain.

Part B - Model Predictive Control (MPC)

• MPC control formulation: simple unconstrained optimal control formulation, general characteristics of MPC formulation, translation of MPC problem into quadratic programming optimisation problem, constant output disturbance observer model, infeasibility and softening of the constraints. 
• Practical MPC implementation considerations: empirical model development, usage of design/tuning parameters, implementation/commissioning of MPC.
• Design of MPC control for typical CSTR chemical reactor as well as distillation column. 

The unit aims to introduce students to the fundamental concepts and their applications of digital control, and the formulation and the main implementation details redarding Model Predictive Control (MPC) as well as the real-time process optimisation.

On successful completion of the course, a student will be able to:

ILO 1: Recognise the relevance of discrete-time models for practical control.

ILO 2: Relate classical control to digital control systems. [Developed and Assessed].

ILO 3: Analyse digital control systems using transfer function and state space modelling techniques. [Developed and Assessed].

ILO 4: Describe Model Predictive Control problem formulation using state-space system model format. [Developed and Assessed].

ILO 5: Convert optimisation-based control problem formulation into general mathematical programming formulation. [Developed and Assessed].

ILO 6: Derive and analyse unconstrained optimal control law for simple low-order single-input, single-output systems. [Developed and Assessed].

ILO 7: Design Model Predictive Control by selecting appropriate weights in the corresponding cost function. [Developed and Assessed].

ILO 8: Summarise the key steps of implementing Model Predictive Control, including development of empirical prediction model and the procedure of controller conditioning. [Developed and Assessed].

The unit covers the following topics:

Part A - Digital Control

• Motivation for digital control theory, including computer-based control.
• Discrete representation of continuous systems: discrete transfer functions, the z transform, and state-space modelling.
• Stability analysis.
• Control system design concerns in practice (e.g. sampling rate selection, computational time) in the discrete domain. 
• Classical anaysis in the discrete domain.

Part B - Model Predictive Control (MPC)

• MPC control formulation: simple unconstrained optimal control formulation, general characteristics of MPC formulation, translation of MPC problem into quadratic programming optimisation problem, constant output disturbance observer model, infeasibility and softening of the constraints. 
• Practical MPC implementation considerations: empirical model development, usage of design/tuning parameters, implementation/commissioning of MPC.
• Design of MPC control for typical CSTR chemical reactor as well as distillation column. 

Lectures and lab sessions.

Exam - 3 hours, 4 questions. Standard feedback is provided after the exam board.

Digital Control - lab. Individual feedback is provided 3 weeks after submission.

MPC Control of Simulated Distillation Column - lab. Online submission, individual performance feedback is provided two weeks after students submit their answers.

Franklin, G. F., Powell, J. D., & Workman, M. L. (1998). Digital Control of Dynamic Systems (3rd ed.). Addison Wesley.

Åström, K. J., & Wittenmark, B. (Eds.). (1984). Computer Controlled Systems: Theory and Design. Prentice-Hall.

Maciejowski, J. M. (2001). Predictive Control with Constraints. Prentice Hall