On Artificial Intelligence for Simulation and Design Space Exploration in Gas Turbine Design

Feature selection and feature learning in machine learning applications for gas turbines: A review

“…Recurrent NNs (RNNs) are widely used for time-series data analysis because they can be used to extract temporal patterns. The applications of ML algorithms for GTs have emerged in the last two decades and have penetrated all aspects of GT development, including diagnostics, prognostics (Fentaye et al, 2019;Tahan et al, 2017), simulation, and design space exploration (Pilarski et al, 2019;Ghalandari et al, 2019). Many projects related to GTs have justified the popularity of DL in this field with superior performance compared with shallow ML algorithms and statistical models (Zhou et al, 2020a;Shen and Khorasani, 2020;.…”