P.D. Domański, J. Arabas, K. Świrski, and M. Wegnerowska (Poland)
Combined Cycle Cogeneration Facility, soft computing
This paper is devoted to the control of a Combined Cycle Cogeneration Facility (CCCF) consisting of two gas turbines (GT), two recovery boilers, and a single steam turbine (ST). The objective of the control is to optimize the distribution of load and steam among the boilers and turbines, taking into account load prediction, equipment efficiency, and equipment availability factors. Three approaches to the dispatch are compared: uniform load distribution (reference method), load dispatch based on gas turbine characteristics only, and a complete economic dispatch also considering the heat recovery boilers performance. Optimal solutions are computed by an Evolutionary Algorithm, whereas the characteristics of the equipment are approximated by Kohonen neural networks. The results of testing the approach for a real power station are provided.
Important Links:
Go Back
