Application concept of artificial neural networks for turbomachinery design. (Received in the final form August 19, 2009)
Journal Title: Computer Assisted Methods in Engineering and Science - Year 2009, Vol 16, Issue 2
Abstract
This paper presents the results of an extensive investigation evaluating and improving the development of artificial neural network (ANN) models for turbomachinery design purposes. A set of 1100 differing axial compressor geometries based on 5 single-stage compressor rigs was prepared. Computations with the mean line analysis tool AXIAL™ took place to determine the according compressor maps defined by 15 operating points each. The challenge of ANN model development in terms of dimensionality reduction (feature selection), data normalization, defining the networks necessary plasticity, and network training is discussed using the example of three different models. As a result, the first model is able to predict the total pressure loss of the rotor blade row with a mean magnitude of the relative error (MMRE) of 3.6%. The second model predicts the total pressure ratio with an average accuracy of 0.8%. The third and last model was trained to predict basic geometrical parameters by presenting the load level and the performance data as an input. The achieved MMRE varied between 2.4% and 5.6% in respect of the particular output variable. The results show that ANNs are applicable to develop efficient models for turbomachinery design and analysis purposes, respectively.
Authors and Affiliations
Mark Azzam, Jan-Christoph Haag, Peter Jeschke
Keywords
Related Articles
- EP ID EP74435
- DOI -
- Views 144
- Downloads 0
Method of fundamental solutions and random numbers for the torsion of bars with multiply connected cross sections
The torsion of bars with multiply connected cross section by means of the method of fundamental solutions (MFS) is considered. Random numbers were used to determine the minimal errors for MFS. Five cases of cross section...
Neural network prediction of load capacity for eccentrically loaded reinforced concrete columns
This paper presents neural networks prediction of load capacity for eccentrically loaded reinforced concrete (RC) columns. The direct modelling of the load capacity of RC columns by means of the finite element method pre...
Identification of electron-phonon coupling factor in a thin metal film subjected to an ultrashort laser pulse
A thin metal film subjected to a laser pulse is considered. The problem is described by the system of energy equations describing the electron gas and lattice temperatures. The thermal interactions between electrons and...
A block sparse shared-memory multifrontal finite element solver for problems of structural mechanics. (Received in the final form July 17, 2009)
The presented method is used in finite-element analysis software developed for multicore and multiprocessor shared-memory computers, or it can be used on single-processor personal computers under the operating systems Wi...
Multiobjective evolutionary optimization of MEMS structures. (Received in the final form November 12, 2010)
The paper is devoted to the shape optimization of piezoelectric and electro-thermo-mechanical devices by the use of multiobjective evolutionary algorithm. In this paper, special implementation of multiobjective evolution...