| Citation: | MA Bowen, WU Xiaoxiong, YU Yang. Surrogate model for deviation angle and total pressure loss prediction of compressor based on machine learning methods[J]. Journal of Aerospace Power, 2023, 38（7）：1675-1690 doi: 10.13224/j.cnki.jasp.20220749
|
In order to improve the prediction accuracy of compressor performance, a surrogate model of deviation angle and total pressure loss based on machine learning methods was built. A two-stage compressor was used as the research object, and the elementary cascade database was established using experimental data of the flow field and the geometric parameters under multiple rotation speed conditions. The sensitivity analysis method was used to screen out the input parameters that have the greatest impact on the deviation angle and total pressure loss. Two machine learning algorithms, i.e.: Gaussian process regression and artificial neural network, were used to establish the deviation angle and total pressure loss model, and Bayesian optimization algorithm was introduced to search for the best hyperparameters of the model. For the optimization iteration and generalization problems faced by artificial neural networks, the Adam algorithm was introduced to adjust the model learning rate and modify the parameter gradient to accelerate the convergence. At the same time, the regularization method was used to enhance the generalization of the model. The cross validation scheme was used in model training process to reduce the risk of overfitting, and the optimal surrogate models were integrated into the throughflow program. The comparison of the calculation results showed, the total pressure ratio prediction error of the surrogate model on low speed conditions was significantly lower than that of the empirical model, among which the artificial neural network modeling had the most significant improvement, and the prediction error was reduced by 0.1 compared with the empirical model. Through comparison of surrogate models, the surrogate model based on artificial neural network had higher prediction accuracy and stronger robustness than the surrogate model based on Gaussian process.
| [1] |
KOCH C,SMITH L. Loss sources and magnitudes in axial-flow compressors[J]. Journal of Engineering for Power,1976,98(3): 411-424. doi: 10.1115/1.3446202
|
| [2] |
CARTER A D S. The low speed performance of related aerofoils in cascade[J]. British NGTE Report No. R.55, 1949.
|
| [3] |
DUNHAM J. Compressor off-design performance prediction using an endwall model[R]. ASME Paper 96-GT-62, 1996.
|
| [4] |
KOENIG M W,HENNECKE D,FOTTNER L. Improved blade profile loss and deviation angle models for advanced transonic compressorbladings: Part Ⅱ a model for supersonic flow[J]. Journal of Turbomachinery,1996,118(1): 81-87. doi: 10.1115/1.2836610
|
| [5] |
BLOCH G, COPENHAVER W, O’BRIEN W. A shock loss model for supersonic compressor cascades[J]. Journal of Turbomachinery, 1999, 121(1): 28-35.
|
| [6] |
徐纲,刘红,李军,等. 一种适用于超跨音叶型的非设计点损失和落后角模型[J]. 航空动力学报,1996,11(1): 4-6.
XU Gang,LIU Hong,LI Jun,et al. A loss and deviation model for transonic axial compressor performance prediction[J]. Journal of Aerospace Power,1996,11(1): 4-6. (in Chinese)
|
| [7] |
SWAN W C. A practical method of predicting transonic-compressor performance[J]. Journal of Engineering for Power,1961,83(3): 322-330. doi: 10.1115/1.3673194
|
| [8] |
刘波, 马乃行, 杨小东, 等. 适应较大叶型弯角范围的轴流压气机落后角模型 [J]. 航空动力学报, 2014, 29(8): 1824-1831
LIU Bo, MA Naixing, YANG Xiaodong, et al. Deviation angle models suitable for wider range of blade profile camber in axial-flow compressor[J]. Journal of Aerospace Power, 2014, 29(8): 1824-1831. (in Chinese)
|
| [9] |
陶慧卿,徐宁,李润泽,等. 基于多元非线性回归损失及落后角模型研究[J]. 热能动力工程,2022,37(12): 38-46.
TAO Huiqing,XU Ning,LI Runze,et al. Research on loss and deviation angle model based on multivariate nonlinear regression[J]. Journal of Engineering for Thermal Energy and Power,2022,37(12): 38-46. (in Chinese)
|
| [10] |
WU Dongrun,TENG Jinfang,QIANG Xiaoqing,et al. Development of an off-design deviation angle prediction model for full blade span in axial flow compressors[J]. Proceedings of the Institution of Mechanical Engineers: Part G Journal of Aerospace Engineering,2019,233(9): 3170-3183. doi: 10.1177/0954410018793265
|
| [11] |
费腾,季路成,周玲. 神经网络模型在压气机通流特性分析中的应用[J]. 航空动力学报,2022,37(6): 1260-1272.
FEI Teng,JI Lucheng,ZHOU Ling. Application of neural network model in compressor through-flow analysis[J]. Journal of Aerospace Power,2022,37(6): 1260-1272. (in Chinese)
|
| [12] |
CUNNAN W S, STEVANS W, URASEK D C. Design and performance of a 427-meter-per-second-tip-speed two-stage fan having a 2.40 pressure ratio[R]. NASA TP-1314, 1978.
|
| [13] |
MORRIS M D. Factorial sampling plans for preliminary computational experiments[J]. Technometrics,1991,33(2): 161-174. doi: 10.1080/00401706.1991.10484804
|
| [14] |
FORRESTER A, SOBESTER A, KEANE A. Engineering design via surrogate modelling: a practical guide[M]. Hoboken, US: John Wiley & Sons, 2008.
|
| [15] |
RASMUSSEN C E. Gaussian processes in machine learning[C]//Proceedings of the Summer School on Machine Learning. Berlin: Springer, 2003: 63-71.
|
| [16] |
HASSOUN M H. Fundamentals of artificial neural networks [M]. Cambridge, US: MIT Press, 1995.
|
| [17] |
TIELEMAN T,HINTON G. Lecture 6.5-rmsprop: divide the gradient by a running average of its recent magnitude[J]. COURSERA:Neural Networks for Machine Learning,2012,4(2): 26-31.
|
| [18] |
LIANG Minfei,LI Zhenming,HE Shan,et al. Stress evolution in restrained GGBFS concrete due to autogenous deformation: bayesian optimization of aging creep[J]. Construction and Building Materials,2022,324: 126690.1-126690.18. doi: 10.1016/j.conbuildmat.2022.126690
|
| [19] |
BERGSTRA J,BENGIO Y. Random search for hyper parameter optimization[J]. Journal of Machine Learning Research,2012,13(2): 281-305.
|
| [20] |
BERGSTRA J, YAMINS D, COX D. Making a science of model search: Hyperparameter optimization in hundreds of dimensions for vision architectures[C]//International Conference on Machine Learning. New York: PMLR, 2013: 115-123. .
|
| [21] |
HUTTER F. Automated configuration of algorithms for solving hard computational problems[D]. Vancouver, Canada: University of British Columbia, 2009.
|
| [22] |
HUTTER F, HOOS H H, LEYTON-BROWN K. Sequential model-based optimization for general algorithm configuration[C]//Proceedings of the International Conference on Learning and Intelligent Optimization. Berlin: Springer, 2011: 507-523..
|
| [23] |
KOHAVI R. A study of cross-validation and bootstrap for accuracy estimation and model selection[C]// International joint conference on Artificial intelligence. Burlington: Morgan Kaufmann Publishers Inc. 1995: 1137-1145.
|
| [24] |
WU C H. A general theory of three-dimensional flow in subsonic and supersonic turbomachines of axial, radial, and mixed-flow types[J]. Transactions of the American Society of Mechanical Engineers,1952,74(8): 1363-1380.
|
| [25] |
NOVAK R A. Streamline curvature computing procedures for fluid-flow problems[J]. Journal of Engineering for Power,1967,89(4): 478-490. doi: 10.1115/1.3616716
|
| [26] |
巫骁雄,刘波,唐天全. 流线曲率法在多级跨声速轴流压气机特性预测中的应用[J]. 推进技术,2017,38(10): 2235-2245. doi: 10.13675/j.cnki.tjjs.2017.10.009
WU Xiaoxiong,LIU Bo,TANG Tianquan. Application of streamline curvature method for multistage transonic axial compressor performance prediction[J]. Journal of Propulsion Technology,2017,38(10): 2235-2245. (in Chinese) doi: 10.13675/j.cnki.tjjs.2017.10.009
|
| [27] |
CREVELING H F, CARMODY R H. Axial-flow compressor computer program for calculating off-design performance[R]. NASA CR-72472.
|
| [28] |
PACHIDIS V. Gas turbine advanced performance simulation[D]. Cranfield, UK: Cranfield University, 2006.
|
| [29] |
BOYER K M,O’BRIEN W F. An improved streamline curvature approach for off-design analysis of transonic axial compression systems[J]. Journal of Turbomachinery,2003,125(3): 475-481. doi: 10.1115/1.1565085
|
DownLoad:
