Uncertainty, modelling, robust predictive control, fuzzy system, process control, energy saving
The research project deals with the development of advanced control methods for systems with uncertainties and focuses on energyintensive processes in the chemical and biotechnologies such as distillation columns, chemical reactors, biochemical reactors, heat exchangers and other processes. The core of the project consists of development of robust predictive control and fuzzy robust control approaches for systems with uncertainties to ensure the control of processes leading to energy savings compared with traditional approaches. Computational effectiveness and usability in practice will be taken into account in the design of control algorithms for systems with uncertainties. Designed algorithms, controllers and control structures will be tested by simulations and laboratory experiments and will be compared with classical control approaches from the viewpoint of energy consumption during the control.
The main objective is to design efficient control structures and algorithms that solve problems related to the presence of uncertainty and of boundaries on manipulated and controlled variables in energy demanding processes representing nodes in chemical, biochemical and food production. Implementation of the control structures and algorithms will result in energy savings in comparison with conventional approaches.
- Design of mathematical models including uncertainties of different types of chemical and biochemical processes;
- Design of fuzzy and neurofuzzy models for processes with uncertainties;
- Extension of the toolbox of mathematical models of chemical and biochemical processes;
- Simulation analysis of processes in terms of their energy consumption;
- Development of robust model predictive control (RMPC) for systems with parametric uncertainties and various boundaries on variables, which guarantees robust stability and required quality of feedback control, and which also reduces conservativeness and the computational burden;
- Transformation of the RMPC design in the form of linear matrix inequalities (LMIs);
- Design of RMPC algorithms reducing computational complexity, and selection of appropriate methods and software tools to solve them,
- Development of RMPC methods using fuzzy and neurofuzzy models of controlled processes with uncertainties,
- Development of a software package for the RMPC design,
- Elaboration of alternative RMPC methods for chemical and biotechnological processes with a focus on the control of energy demanding processes (heat exchangers, distillation columns, chemical and biochemical reactors, crystallizers),
- Comparison of the proposed RMPC approaches with classical approaches based on simulation results.