EXERGY DIAGNOSIS OF COAL FIRED COMBINED HEAT AND POWER PLANT WITH APPLICATION OF NEURAL AND REGRESSION MODELLING by

Mathematical models of the processes, that proceed in energetic machines and devices, in many cases are very complicated. In such cases, the exact analytical models should be equipped with the auxiliary empirical models that describe those parameters which are difficult to model in a theoretical way. Regression or neural models identified basing on measurements are rather simple and are cha-racterized by relatively short computation time. For this reason they can be effec-tively applied for simulation and optimization of steering and regulation processes, as well as, for control and thermal diagnosis of operation (e. q. power plants or combined heat and power plants). In the paper regression and neural models of thermal processes developed for systems of operation control of ther-mal plants are presented. Theoretical-empirical model of processes proceeding in coal fired combined heat and power plant have been applied. Simulative calcu-lations basing on these models have been carried out. Results of simulative calcu-lations have been used for the exergetic evaluation of considered power plant. The diagnosis procedure let to investigate the formation of exergy costs in inter-connected components of the system of combined heat and power, as well as, in-vestigate the influence of defects in operation of components on exergy losses and on the exergetic cost in other components. Key words: mathematical modelling, neural networks, combined heat and power plant, exergy, diagnosi