Abstract The turbine, as a critical component of gas turbines, has been widely used in marine propulsion. However, under small-sample data conditions, it is confronted with many challenges like calibration difficulties and over-fitting. To address these issues, a rapid performance prediction method based on loss model theory and small-sample data-driven adaptive calibration was proposed. Various energy loss models were integrated and the optimal model was identified through systematic multi-criteria evaluation. Key parameters were determined using Self-Organizing Map analysis for intelligent dimensionality reduction. This process identified the most sensitive coefficients, which reduced the number of parameters from 29 to 8 and cut calibration time by 50%. For the single-stage turbine, the maximum prediction error was reduced from 3.84% to 0.83%. High accuracy was maintained in multi-stage turbines, where inherent overestimation of losses was effectively corrected as validated by 3D flow details. Practical value was demonstrated by embedding the model into a zero-dimensional dynamic simulation with a maximum relative error of 1.0%. This research supports the construction and optimization of digital models for gas turbines.
Zheng et al. (Fri,) studied this question.