Læringsutbytte

The course will provide learning outcomes in terms of knowledge, skills and general competence as follows:

Knowledge:

• The students will learn the theoretical basis for model-based simulation and control design of industrial processes.

Skills:

• The students will be able to design, simulate and implement control systems on selected software platforms.
• The students will be able to test different control methods and control strategies using dynamic simulators.

General competences:

• The students will be able to assess the usefulness of dynamic simulation and control design with regard to technology, environment and economy.
• The students will be able to understand literature and communicate with professionals about model-based control design of industrial processes.
• The students will be able to document and disseminate the results of projects on model based control design.

Innhold

Mathematical modeling: from linear ordinary differential equations to transfer functions. Linearization of nonlinear systems using Taylor series and deviation variables. Model parameter fitting using least squares techniques. Control Systems Diagrams, Instrumentation diagrams, block diagrams and signal flow graphs. Use of Mason’s gain rule to get the transfer function of complex systems. Representation of MIMO systems using transfer functions matrices. Discussion of dead-time. Discussion of common non-linear characteristics: saturation, hysteresis, dead-band, backlash. Implementation of realistic simulators for testing of control methods and control strategies, including simulation based criteria for assessing control performance. Introduction to the analysis of dynamic systems by using the location of zeros and poles. Effect of the system poles on the stability, oscillations and settling time of the system. Effect of the system zeros on the dynamic response of the system: analysis and simulation of systems with inverse response. Stability analysis using frequency response. Introduction to optimization: optimization criteria, variables, constraints. Setting and solving control related optimization problems using different software tools. Introduction to state space representation. State space realization from ordinary differential equations. Minimal state space realizations from transfer functions using matlab. Controllability, observability and stability analysis using state space realizations in matlab. Introduction to observers and the Kalman filter. Introduction to model predictive control. Implementation of the algorithms using matlab/simulinik.

Arbeids- og læringsformer

The teaching methodology will use lectures in two sessions per week. During the first session theory and methods will be presented and demonstrated with simple systems. During the second session extensive simulation exercises and application of the topics developed during the first session will be applied to more complex systems (an helicopter system and a four tanks level system) . There will be both individual and group-based learning activities.

Vurderingsformer

A course grade (A-F) will be set on the basis of an individual written final exam with a value of 70% of the course grade, and a project with a value of 30% of course grade. There will also be a number of mandatory exercises that the students must complete before the final exam. The instructors can assign the project individually or in teams. The evaluation of the project will be based both in a report and an oral presentation by the students. The students must get a passing grade in both the final exam and the project in order to pass the course.

Det tas forbehold om mindre justeringer i planen.