This paper presents three different gas turbine condition monitoring models developed by artificial neural networks. Operational data from an ALSTOM GT11-N1 has been employed for training and evaluation of the artificial neural network models. The developed models differ by their selected input and output parameters. These ANN models are used for prediction of performance parameters of the gas turbine. When tested on data collected after a training period, the models reveal different degradation trends, i.e. difference between expected and measured performance parameter values. The approach adopted in this paper shows that powerful monitoring tools can be developed form operational data with artificial neural networks.

This content is only available via PDF.
You do not currently have access to this content.