Since the late 1930s, gas turbine has begun to develop rapidly. To improve the economic and safety of gas turbine, new types were generated frequently by Original Equipment Manufacture (OEM). In this paper, a hybrid GRA-SVM prediction model is established to predict the main design parameters of new type gas turbines, based on the combination of Grey Relational Analysis (GRA) and Support Vector Machine (SVM). The parameters are classified into two types, system performance parameters reflecting market demands and technology development, and component performance parameters reflecting technology development and coupling connections. The regularity based on GRA determines the prediction order, then new type gas turbine parameters can be predicted with known system parameters. The model is verified by the application to SGT600. In this way, the evolution rule can be obtained with the development of gas turbine technology, and the improvement potential of several components can be predicted which will provide supports for overall performance design.
- International Gas Turbine Institute
Design Parameters Prediction of New Type Gas Turbine Based on a Hybrid GRA-SVM Prediction Model
- Views Icon Views
- Share Icon Share
- Search Site
Wei, T, Zhou, D, Chen, J, Cui, Y, & Zhang, H. "Design Parameters Prediction of New Type Gas Turbine Based on a Hybrid GRA-SVM Prediction Model." Proceedings of the ASME Turbo Expo 2017: Turbomachinery Technical Conference and Exposition. Volume 3: Coal, Biomass and Alternative Fuels; Cycle Innovations; Electric Power; Industrial and Cogeneration Applications; Organic Rankine Cycle Power Systems. Charlotte, North Carolina, USA. June 26–30, 2017. V003T06A014. ASME. https://doi.org/10.1115/GT2017-63905
Download citation file: