W artykule przedstawiono wyniki badań identyfikacyjnych wielowymiarowego modelu paleniska fluidalnego, przedstawionego w formie transmitancji macierzowej, zdyskretyzowanej do postaci ARX. Zaproponowano warunki przebiegu eksperymentu zapewniającego identyfikowalność modelu dynamiki, przy ograniczeniu ryzyka awaryjnego wyłączenia kotła. Na podstawie analizy działania paleniska fluidalnego wskazano typowe zakłócenia procesu spalania z uwzględnieniem podziału na zakłócenia mierzalne i niemierzalne oraz sformułowano metodę identyfikacji wpływu tych zakłóceń na sygnały wyjściowe modelu. Wskazano możliwość aplikacji przedstawionego modelu w regulatorach bazujących na modelu automatyzowanego procesu, np. predykcyjnych, adaptacyjnych czy IMC (Internal Model Control).
EN
The paper presents results of the identification of the multivariable model of the fluidized bed boiler derived in the form of transfer function matrix. Conditions of the experiment that ensure the identifiability of model dynamics while limiting the risk of emergency boiler shutdown was proposed. Typical disturbances of the combustion process including division into measurable and immeasurable ones as well as the method to identify the impact of these disturbances on the outputs of the model was formulated. Possible application of the model in the model-based controllers such as predictive or adaptive IMC (Internal Model Control) was indicated.
The paper presents proposal of a model of the fluidized bed boiler adapted for use in model-based controllers e.g. predictive, adaptive or internal model control (IMC). The model has been derived in the form of transfer function matrix which allows its direct implementation in the controller structure. Formulated model takes into consideration the principal cross-coupling between process variables which enables the opportunity to search for feasibility of decoupling control. The results of the identification of the dynamics of the 2 MW industrial bubbling fluidized bed boiler using the proposed model form was presented. According to the experimental data it was found that despite of introduced simplifications presented model allows the boiler behavior prediction.
This paper presents a method to verify the controllability and observability of linear multivariable systems with multiple modes using eigenprojectors or constituent matrices. Eigenprojectors are obtained by the partial fraction expansion of the numerator of the transfer function matrix of the system and the constituent matrices are obtained from the spectral decomposition of the state matrix. Eigenprojectors inform on the properties of the corresponding mode separately from the others, so controllability and observability of each mode are checked. A previous paper has dealt with linear systems with single eigenvalues, and the same method gave good results. Numerical examples are given to illustrate this method.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.