Predictive, adaptive model of PG 9171E gas turbine unit including control algorithms

Contemporary thermal diagnostic systems require computational tools, including mathematical models. Because of required short computation time, these models should have a simple structure. Therefore very often, for these purposes, analytical-empirical models are used. Such models encompass both mass and energy balances and additional empirical functions whose coefficients are estimated by using the measurement results. As a result, changing technical conditions are taken into account. This paper presents a simulation model of PG 9171E gas turbine unit by General Electric which contains partial models of an axial compressor, low-emission combustion chambers, and an axial expander. The unknown values of empirical coefficients were estimated based on the operating data using the least squares method. The simulation model allows the calculation of non-measured operating parameters and energy assessment indicators, e.g. efficiency of electricity production. An important advantage of the developed model is that it has the capability of adapting to the changing technical conditions of the machine. The results of calculations were compared to the results of measurements. Model predictive quality was verified with the use of the determination factor and root mean square error.