Title :
Application of Artificial Neural Networks in turbomachinery optimization
Author :
Jin, Cai ; Yuan, Li ; Jian-feng, Yu ; Zheng, Wang
Author_Institution :
Minist. of Educ. Key Lab. of Contemporary, Design & Integrated Manuf. Technol., Northwestern Polytech. Univ., Xi´´an, China
Abstract :
The Response surface methodology provides a way to engineering optimization problems of the turbomachinery. Using a prototypical Artificial Neural Networks (ANN) problem, turbomachinery aerodynamic optimization can improve the efficiency and performance. This study explores the approximation of the turbomachinery objective function which has a significant impact on the accuracy and effectiveness of the resulting ANN model. Application of straightforward simulation computation time improved the generalization capability of the ANN model, avoiding the potential problems of over-training or memorization meanwhile.
Keywords :
aerodynamics; mechanical engineering computing; neural nets; optimisation; response surface methodology; turbomachinery; artificial neural network; function approximation; generalization capability; simulation computation; turbomachinery aerodynamic optimization; Approximation methods; Artificial neural networks; Neurons; Optimization; Surface reconstruction; Surface topography; ANN (Artificial Neural Networks); Aerodynamic Optimization; Turbomachinery;
Conference_Titel :
Intelligent Computing and Intelligent Systems (ICIS), 2010 IEEE International Conference on
Conference_Location :
Xiamen
Print_ISBN :
978-1-4244-6582-8
DOI :
10.1109/ICICISYS.2010.5658319
