Gas path model identification as an instrument of gas turbine diagnosing

This paper deals with a development of identification procedures of gaspath static and dynamic models and with their application for gas turbine diagnosing. The identification is important because the parameters estimated during this task aid to detect particular degraded components. The first purpose of our work was to develop and test the identification procedure of a nonlinear dynamic model. A statistical procedure testing on the model input data was applied to analyze diagnostic possibilities of the developed procedure. The second purpose was to present and generalize previous investigations related with static model identification and to give any common approach of both techniques application in gas turbine diagnostics.