Comparison of various regression models for predicting compressor and turbine performance parameters

This paper investigates various regression models to predict the compressor and turbine map parameters. To this end, we used the data measured from two different types of compressors and turbines. The compressor is basically a single stage radial machine, whereas, the turbine is formed of high pressure and low pressure parts. The emphasis of this study is to construct various models for prediction of corrected mass flow rate and isentropic efficiency. Except for prediction capabilities, the study also compares the regression model’s performances. Results show that the designed models can be used for the development of dynamic mathematical model of a gas turbine engine. •Regression models are constructed for prediction of the efficiency and mass flow rate.•Models are compared in terms of the calculated coefficients’ p and AIC values.•Linear and the most popular nonlinear regression techniques are investigated.•The constructed models can provide a mathematical basis for engine design studies.•The models can be adapted to various engine types easily.