Artificial Neural Network Modeling of Unsteady Aerodynamic Characteristics of Aircraft at High Attack Angle
-
摘要: 大攻角气动特性预测与气动建模是新型飞行器提升飞行性能的重要内容.以轴对称导弹简化模型为研究对象,首先采用计算流体力学方法,对70°大攻角状态的非定常气动特性进行数值模拟,计算方法基于RANS的N-S方程,湍流模型采用SA模型,对流场采用有限体积法离散,无黏项采用Roe通量差分分裂格式,黏性项采用中心差分,时间推进采用LU-SGS格式的双时间步法.飞行器运动模式采用强迫振荡的方式,对5种不同振荡频率进行了非定常数值计算,并记录每一内迭代周期最终的气动力和力矩数值.其次,以CFD预测结果作为气动建模的样本,采用动导数模型、多项式模型等传统方法,进行气动建模,并分析其有效性和精度.最后采用神经网络方法对大攻角非定常气动力进行建模,并和动导数模型、多项式模型进行精度对比.结果表明,基于神经网络的人工智能气动建模方法具有较高的精度和适应性.该方法为飞行器大攻角非定常非线性气动建模,大攻角飞行稳定性分析与控制提供理论参考.Abstract: The simulation and modeling of aerodynamic characteristics at high angle of attack is one of the most important research areas for new concept aircraft. Based on simplified missile model, firstly, aerodynamic characteristics of 70å ttack angle were numerically simulated by RANS-based CFD method with SA turbulence model. The finite volume method was used to discretize the N-S formulation. The LU-SGS dual time-stepping algorithm was used for time marching. The unsteady calculations with five different oscillation frequencies were carried out in the mode of forced oscillation, and the final aerodynamic data in each iteration period were recorded. Secondly, based on CFD results, traditional methods such as dynamic derivative model and polynomial model were used for aerodynamic modeling, and the validity and accuracy were analyzed. Finally, dynamic derivative model, polynomial model and neural network mode were used to modeling aerodynamic characteristics. The results show that the artificial intelligence aerodynamic modeling method based on neural network has higher accuracy and adaptability. This method provides theoretical technical support for the unsteady nonlinear aerodynamic modeling, and the stability analysis and control of aircrafts at high angle of attack.
-
表 1 动导数建模结果
Table 1. Aerodynamic derivative modeling results
samples Cmz0 Cmzα Cmzωz 1 Hz -0.993 5 -0.001 0 -1.935 6 2 Hz -0.995 0 -0.000 9 -1.937 4 3 Hz -1.002 9 -0.001 0 -1.953 2 5 Hz -0.992 9 -0.000 8 -2.001 5 10 Hz -0.993 5 -0.001 0 -1.935 6 表 2 3阶多项式建模结果
Table 2. 3rd order polynomial modeling results
samples Cmz0 Cmzα Cmzωz 1 Hz -0.926 7 -0.000 5 -1.951 6 2 Hz -0.926 0 -0.000 7 -1.953 7 3 Hz -0.930 7 -0.001 0 -1.955 5 5 Hz -0.961 9 -0.000 8 -2.004 5 10 Hz -0.858 3 -0.000 6 -2.066 8 表 3 非定常气动建模精度对比
Table 3. Comparison of unsteady aerodynamic modeling accuracy
samples dynamic derivatives model/% 3rd polynomial model/% neural networks model/% 1 Hz 9.67 1.3 0.47 2 Hz 10.02 1.17 0.49 3 Hz 12.09 1.57 0.42 5 Hz 12.44 1.04 0.52 10 Hz 9.67 1.04 0.46 all data 10.1 1.6 0.5 -
[1] 汪清, 钱炜祺, 丁娣.飞机大迎角非定常气动力建模研究进展[J].航空学报, 2016, 37(8):2331-2347. http://d.old.wanfangdata.com.cn/Periodical/hkxb201608002Wang Q, Qian W Q, Ding D. A review of unsteady aerodynamic modeling of aircrafts at high angles of attack[J]. Acta Aeronautica et Astronautica Sinica, 2016, 37(8):2331-2347(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/hkxb201608002 [2] 袁先旭, 张涵信, 谢昱飞.基于CFD方法的俯仰静、动导数数值计算[J].空气动力学学报, 2005, 23(4):458-463. doi: 10.3969/j.issn.0258-1825.2005.04.012Yuan X X, Zhang H X, Xie Y F. The pitching static/dynamic derivatives computation based on CFD methods[J]. Acta Aerodynamica Sinica, 2005, 23(4):458-463(in Chinese). doi: 10.3969/j.issn.0258-1825.2005.04.012 [3] 刘绪, 刘伟, 柴振霞, 等.飞行器动态稳定性参数计算方法研究进展[J].航空学报, 2016, 37(8):2348-2369. http://d.old.wanfangdata.com.cn/Periodical/hkxb201608003Liu X, Liu W, Chai Z X, et al. Research progress of numerical method of dynamic stability derivatives of aircraft[J]. Acta Aeronautica et Astronautica Sinica, 2016, 37(8):2348-2369(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/hkxb201608003 [4] Wang F J. Numerical prediction of stability derivatives for complex configurations[R]. APISAT2014, 2014. [5] 陈农.大攻角非定常气动力建模研究[D].北京: 中国航天空气动力技术研究院, 2007.Chen N. High angles of attack unsteady aerodynamic modeling[D]. Beijing: China Academy of Aerospace Aerodynamics, 2007(in Chinese). [6] 黄达.飞行器大振幅运动非定常空气动力特性研究[D].南京: 南京航空航天大学, 2007. http://cdmd.cnki.com.cn/article/cdmd-10287-2007193997.htmHuang D. Unsteady aerodynamic characteristics for the aircraft oscilation in large amplitude[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2007(in Chinese). http://cdmd.cnki.com.cn/article/cdmd-10287-2007193997.htm [7] 蔡金狮.飞行器系统辨识[M].宇航出版社, 1995.Cai J S. System identification for flight vehicles[M]. China Astronautic Publishing House, 1995(in Chinese). [8] 王贵东.航天飞行器气动参数辨识研究[D].中国航天空气动力技术研究院, 2010.Wang G D. Study on aerodynamic parameter identification for aerospace vehicle[D]. China Academy of Aerospace Aerodynamics, 2010(in Chinese). [9] Lin G F, Lan C E. A Generalized dynamic aerodynamic coefficient model for flight dynamics application[R]. AIAA-97-3643, 1997. [10] Allwine D A, Strhaler J A, Lawernee D A, et al. Nonlinear modeling of unsteady aerodynamics at high angle of attack[R]. AIAA 2004-5275, 2004. [11] Goman M, Khrabrov A. State-space representation of aerodynamic characteristics of an aircraft at high angles of attack[R]. AIAA 1992-4651, 1992. [12] 陈海萍, 高正红.大迎角非定常气动力数学模型研究[J].飞行力学, 2008, 26(5):10-12, 16. http://d.old.wanfangdata.com.cn/Periodical/xbgydxxb200104004Chen H P, Gao Z H. Large angle of attack unsteady aerodynamic modeling study[J]. Flight Dynamics 2008, 26(5):10-12, 16(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/xbgydxxb200104004 [13] 刘志涛, 孙海生, 姜裕标, 等.非线性非定常气动力的模糊逻辑建模方法[J].实验流体力学, 2005, 19(1):99-103. doi: 10.3969/j.issn.1672-9897.2005.01.020Liu Z T, Sun H S, Jiang Y B, et al. Fuzzy logic modeling of nonlinear unsteady aerodynamics[J]. Journal of Experiments in Fluid Mechanics, 2005, 19(1):99-103(in Chinese). doi: 10.3969/j.issn.1672-9897.2005.01.020 [14] 吴辰, 姚宏, 彭兴钊, 等.支持向量回归机在飞机气动力建模中的应用[J].计算机仿真, 2013, 30(10):128-132. doi: 10.3969/j.issn.1006-9348.2013.10.030Wu C, Yao H, Peng X Z, et al. Application of support vector regression for aerodynamic modeling[J]. Computer Simulation, 2013, 30(10):128-132(in Chinese). doi: 10.3969/j.issn.1006-9348.2013.10.030 [15] Chan Y Y, Zhu Z W. Neural network modeling of aerodynamic[R]. AIAA 99-0685, 1999. [16] Steck J, Rokhsaz K. Some applications of artificial neural networks in modeling of nonlinear aerodynamics and flight dynamics[R]. AIAA 1997-0338, 1997. [17] 陈海, 钱炜祺, 何磊.基于深度学习的翼型气动系数预测[J].空气动力学学报, 2018, 36(2):294-299. doi: 10.7638/kqdlxxb-2017.0098Chen H, Qian W Q, He L. Aerodynamic coeffient prediction of airfoils based on deep learning[J]. Acta Aerodynamica Sinica, 2018, 36(2):294-299(in Chinese). doi: 10.7638/kqdlxxb-2017.0098 [18] 张瑞民, 张石玉, 赵俊波.基于神经网络的非定常气动力建模研究[J].计算机仿真, 2017, 34(2):106-110. doi: 10.3969/j.issn.1006-9348.2017.02.024Zhang R M, Zhang S Y, Zhao J B. The research of neural network in modeling of unsteady aerodynamics[J]. Computer Simulation, 2017, 34(2):106-110(in Chinese). doi: 10.3969/j.issn.1006-9348.2017.02.024 [19] 龚正, 沈宏良.非定常气动力的结构自适应神经网络建模方法[J].飞行力学, 2007, 25(4):13-16. doi: 10.3969/j.issn.1002-0853.2007.04.004Gong Z, Shen H L. Structure self-adapting ANN method in modeling of unsteady aerodynamic[J]. Flight Dyna-mics, 2007, 25(4):13-16(in Chinese). doi: 10.3969/j.issn.1002-0853.2007.04.004 [20] 王博斌, 张伟伟, 叶正寅.基于神经网络模型的动态非线性气动力辨识方法[J].航空学报, 2010, 31(7):1379-1388. http://d.old.wanfangdata.com.cn/Periodical/hkxb201007011Wang B B, Zhang W W, Ye Z Y. Unsteady nonlinear aerodynamics identification based on neural network model[J]. Acta Aeronautica et Astronautica Sinica, 2010, 31(7):1379-1388(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/hkxb201007011 [21] 史志伟, 王峥华, 李俊成.径向基神经网络在非线性非定常气动力建模中的应用研究[J].空气动力学学报, 2012, 30(1):108-119. doi: 10.3969/j.issn.0258-1825.2012.01.019Shi Z W, Wang Z H, Li J C. Numerical study of flow characteristics of a plunging rigid airfoil with elastic trailing-edge plate[J]. Acta Aerodynamic Sinica, 2012, 30(1):108-119(in Chinese). doi: 10.3969/j.issn.0258-1825.2012.01.019 [22] 付军泉, 史志伟, 陈坤, 等.基于EKF的实时循环神经网络在非定常气动力建模中的应用[J].空气动力学学报, 2018, 36(4):658-663. doi: 10.7638/kqdlxxb-2016.0131Fu J Q, Shi Z W, Chen K, et al. Applications of real-time recurrent neuralnetwork based on extended Kalman filter in unsteady aerodynamics modeling[J]. Acta Aero-dynamic Sinica, 2018, 36(4):658-663(in Chinese). doi: 10.7638/kqdlxxb-2016.0131 [23] 王刚, 刘钧圣, 王琨, 等.一种亚声速导弹气动力计算方法[J].弹箭与制导学报, 2018, 38(2):65-68. http://d.old.wanfangdata.com.cn/Periodical/djyzdxb201802015Wang G, Liu J S, Wang K, et al. A method to predict aerodynamic characteristics for subsonic missile[J]. Journal of Projectiles, Rockets, Missiles and Guidance, 2018, 38(2):65-68(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/djyzdxb201802015 [24] 吴子牛.计算流体力学基本原理[M].北京:科学出版社, 2001.Wu Z N. The basic principles of computational fluid mechanics[M]. Beijing:Science Press, 2001(in Chinese). [25] 杨云军.飞行器非稳定运动的流动物理及动力学机理[D].北京: 中国航天空气动力技术研究院, 2008.Yang Y J. Flow physics and dynamic mechanism of non-stable motion of aircraft[D]. Beijing: China Academic of Aerospace Aerodynamics, 2008(in Chinese). [26] 张瑞民, 时晓天.有翼导弹动态气动特性数值研究[J].弹箭与制导学报, 2017, 37(1):117-120, 128. http://d.old.wanfangdata.com.cn/Periodical/djyzdxb201701027Zhang R M, Shi X T. Research on numerical virtual flight of spinning projectile[J]. Journal of Projectiles, Roc-kets, Missiles and Guidance, 2017, 37(1):117-120, 128(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/djyzdxb201701027 [27] 孙海生, 张海酉, 刘志涛.大迎角非定常气动力建模方法研究[J].空气动力学报, 2011, 29(6):733-737. http://d.old.wanfangdata.com.cn/Periodical/kqdlxxb201106008Sun H S, Zhang H Y, Liu Z T. Comparative evaluation of unsteady aerodynamics modeling approaches at high angle of attack[J]. Acta Aerodynamica Sinica, 2011, 29(6):733-737(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/kqdlxxb201106008 [28] 史志伟, 吴根兴.大攻角非定常气动力建模与模型比较[J].空气动力学学报, 1999, 17(4):454-461.Shi Z W, Wu G X. Comparison between the modeling and model of the non-fixed normal gas power of the large attack angle[J]. Acta Aerodynamica Sinica, 1999, 17(4):454-461(in Chinese). [29] Bryan G H. Stability in aviation[M]. Lonton:Macmil-lan and Co., 1911:20-105. [30] 沈霖.大攻角非定常气动力建模及尾旋仿真研究[D].南京: 南京航空航天大学, 2013.Shen L. Research on unsteady aerodynamic models at high angle of attack and spin simulation of aircraft[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2013(in Chinese). [31] 白斌, 徐敏, 祝小平, 等.正交多项式在非定常气动建模上的运用[J].飞行力学, 2013, 31(5):398-401. doi: 10.3969/j.issn.1002-0853.2013.05.004Bai B, Xu M, Zhu X P, et al. Nonlinear aerodynamic modeling using on orthogonal polynomials[J]. Flight Dynamics, 2013, 31(5):398-401(in Chinese). doi: 10.3969/j.issn.1002-0853.2013.05.004 [32] 王超, 王贵东, 白鹏.飞行仿真气动力数据机器学习建模方法[J].空气动力学学报, 2019, 37(3):488-497. doi: 10.7638/kqdlxxb-2019.0024Wang C, Wang G D, Bai P. Method of flight test aerodynamic modeling based on machine learning[J]. Acta Aerodynamic Sinica, 2019, 37(3):488-497(in Chinese). doi: 10.7638/kqdlxxb-2019.0024 [33] 赵忠良, 任斌, 黄叙辉. SDM标模大攻角动导数试验[J].航空学报, 1998, 19(2):137-141. http://www.cnki.com.cn/Article/CJFDTotal-HKXB802.001.htmZhao Z L, Ren B, Huang X H. High angle of attack dynamic derivative experiment of standard dynamic model[J]. Acta Aerodynamica Sinica, 1998, 19(2):137-141(in Chinese). http://www.cnki.com.cn/Article/CJFDTotal-HKXB802.001.htm [34] 钱杏芳, 林瑞雄, 赵亚男.导弹飞行力学[M].北京:北京理工大学出版社, 2000.Qian X F, Lin R X, Zhao Y N. Missile flight dyna-mics[M]. Beijing:Beijing Institute of technology, 2000(in Chinese).