In the paper, the research on reversed modeling technology and its application with operation data of power plant is presented with the consideration of nonlinear and multiple coupling characteristics of thermal system in large capacity of power plant. The paper focus on the partial least-squares regression and artificial neural networks method which were studied based on data-driven model. The PLS-ANN coupling model were proposed. The paper aimed at the problems occurred in the operation of power generating units. Finally, the unburned carbon in fly ash model based on PLS regression and ANN method were built under the consideration of the advantages of the above models. The simulation results indicated that the presented coupling model are good at the model of non-linear system with the characteristics of higher computing speed as well as high precision. By presented model, the analysis of the impact of the unburned carbon in fly ash could be carried out in consideration of the changes of operating parameters as well as coal quality. The model could also be expected to be used in guiding of the optimal operation of power plant boiler.
|