Development of surrogate models for estimating transient temperature of hot gas path part

Abstract

To develop a real-time temperature estimating method for a gas turbine blade and a header of heat recovery steam generator, surrogate models that can estimate transient changes in temperature under arbitrary operation conditions were obtained by applying Gaussian process regression to results of numerical simulations. The predicted temperatures by the surrogate model adequately agreed with the numerical results under various operating conditions. The surrogate model was found to be robust against the noise in operating condition data and setting error of initial temperature. In addition, the surrogate model for predicting thermal stress was also obtained and accuracy of the model was validated.