Návrh a implementace prediktivního řízení

Abstract

This doctoral thesis is focused on the superheated steam temperature control in a biomass power plant. The theoretical bases of work are model predictive control design and description of superheated steam temperature technology. The first practical verification of designed model predictive control algorithm was performed in the laboratory on the hot air laboratory model. The control algorithm was realized by using software MATLAB. The main focus of the thesis was to design sophisticated model predictive control for superheating technology in the high pressure part of the drum boiler that meets the technological requirements in the entire power range of the installed drum boiler and also implement it into a selected hardware platform by applying C++ programming language. In accordance with the aim of the thesis an optimal control structure for the specified technological modes and power levels was found. In the thesis two control structures are presented. The first realized control was based on conventional PI controllers in cascade structure. The second control structure represents model predictive control. In addition, both approaches were simulated in software MATLAB/Simulink. At the end of the thesis the verification of the control algorithms in real operation in the biomass power plant is performed. The original installed superheated steam temperature control is compared with the newly designed cascade control and subsequently cascade control with model predictive control. The completely new designed superheated steam temperature control was elaborated with its subsequent implementation into the existing control system in the biomass power plant.