针对传统试验和仿真手段获取多工况复杂流场信息成本高、耗时长问题,提出了一种基于特征代理模型的复杂流场预测方法。在有限个样本流场的基础上采用本征正交分解获得包含流场特征的基模态流场,在此基础上采用多维变量径向基插值实现新流场预测。结果表明,该方法可以对复杂结构流场进行预测,在亚声速、跨声速和超声速等不同马赫数阶段都具有较好的适用性,并在亚声速和超声速条件下具有更高的预测精度。
A new method based on characteristic surrogate model for prediction of complicated flow fields was proposed, due to the problem of high cost and long time consuming of complex flow field information obtained by traditional test and simulation method. Proper orthogonal decomposition was selected to acquire basis model flow fields with flow field characteristics based on finite set of sample flow fields, and on this basis, multidimensional variable radial basis function interpolation was used for the prediction of new flow fields. The results show that this method can predict the flow field with complex structures, and it has good applicability in different Mach number stages, such as subsonic speed, transonic speed and supersonic speed. Meanwhile, it has higher prediction accuracy in subsonic and supersonic conditions.
[an error occurred while processing this directive]
[1] BELSON B A, TU J H, ROWLEY C W. Algorithm 945: modred:a parallelized model reduction library[J].ACM Transactions on Mathematical Software, 2014, 40(4): 1-23.
[2] LUMLEY J L. The structure of inhomogeneous turbulence in atmospheric turbulence and wave propagation[M]//TATARSKI V I. Wave Propagation in a Turbulent Medium. Moscow:[s.n.], 1961: 166-178.
[3] HOLMES P J, LUMLEY J L, BERKOOZ G, et al. Turbulence, coherent structures, dynamical systems and symmetry[M].2nd ed. Cambridge: Cambridge University Press, 2011: 86-113.
[4] SIMPSON T W, BOOKER A J, GHOSH D, et al. Approximation Methods in Multidisciplinary Analysis and Optimization: a panel discussion[J].Structural and Multidisplinary Optimization, 2004, 27(5): 303-313.
[5] WANG A X, MA Y C, YAN W J. The proper orthogonal decomposition method for Navier-Stokes equations[J].Academic Journal of Xi’an Jiaotong University, 2008, 20(3): 141-148.
[6] 邱亚松, 白俊强, 华俊. 基于本征正交分解和代理模型的流场预测方法[J].航空学报, 2013, 34(6): 1249-1260.