Model-based process optimisation

In control engineering the key methods for analysing and designing control systems are based on the model of the process or the system. These models are typically derived from physical considerations and master equations and then adapted to the existing machine using practical measurements.

Core Area

Every process can be optimised, e.g. with regard to operating costs and specifically with respect to energy efficiency. The Institute of Energy Systems and Fluid Engineering (IEFE) does this in the area of model-based process optimisation, using a system model with an optimal automation concept.

The design for the practical implementation of an automation concept considering the model-based process optimisation includes the elements shown here (see picture 1).

Expertise

Metrology

Metrology is a very important step in model-based process optimisation, since recording the static and dynamic process parameters alone says a lot about the behaviour and the energy use of the system. Furthermore, metrology is required to adapt the theoretical model to the existing system.

When implementing an automation concept, metrology also includes providing all the sensor signals for the respective downstream control. The software LabVIEW is used to record, process, save and depict process factors.

 

Modelling and simulation

Simulink-Model to determine fluid speed in real-time. A mathematical model of a test rig is being deduced from energy-accounting equations on the basis of physical examinations.

The so-called top-down approach allows the initial model to be derived as simple as possible to limit modelling costs. The model should only become more complex when necessary.

In practice, a mathematical model of a system cannot always be derived, or is too complex and too costly. In these cases the system is, for example, operated with test signals for experimental identification of the respective system model. IEFE uses the MATLAB or Simulink software for modelling, simulation and identification purposes.

Real time implementation of automation concepts

The designed automation concept is first verified with simulation models by using system models.

The goal is to test as many aspects of the automation concept as possible before performing test runs with the existing system. This allows IEFE to save operating costs in the prototype phase.

Changes or additions to the automation concept can easily be implemented with hardware in the loop or software in the loop during this development phase and prior to the final implementation of the process control.

Real-time implementation

Model-based process optimisation project examples