Multi-kernel neural networks for nonlinear unsteady aerodynamic reduced-order modeling

Kou, Jiaqing; Zhang, Weiwei

doi:10.1016/j.ast.2017.04.017

Cited by 54 publications

(19 citation statements)

References 83 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…limit cycle oscillation under the nonlinearity of a large-amplitude structural motion or flow separation, practical nonlinear ROMs have been developed, such as the Volterra series model [80], the neural network model [81], and the Winer model [82]. To improve the generalization ability of these nonlinear ROMs and construct linear/non-linear combination models, Kou [83][84] and Winter [85][86][87] have carried out extensive original researches. In recent years, increasing attention has been paid to deep learning and machine learning methods in this field [88].…”

Section: Reduced-order Model For the Unsteady Flowmentioning

confidence: 99%

Transonic aeroelasticity: A new perspective from the fluid mode

Gao

Zhang

2020

Progress in Aerospace Sciences

Self Cite

View full text Add to dashboard Cite

Within the transonic regime, the aeroelastic problems exhibit many unique characteristics compared with subsonic and supersonic cases. Although a lot of research has been carried out in this field, the underlying mechanisms of these complex phenomena are not clearly understood yet, resulting in a challenge in the design and use of modern aircraft. This review summarizes the recent investigations on nonclassical transonic aeroelastic problems, including transonic buzz, reduction of transonic buffet onset, transonic buffeting response and frequency lock-in phenomenon in transonic buffet flow. After introducing the research methods in unsteady aerodynamics and aeroelastic problems, the dynamical characteristics as well as the physical mechanisms of these phenomena are discussed from the perspective of the fluid mode. In the framework of the ROM (reduced order model) -based model, the dominant fluid mode (or the eigenvalue) and its coupling process with the structural model can be clearly captured. The flow nonlinearity was believed to be the cause of the complexity of transonic aeroelasticity. In fact, this review indicates that the complexity lies in the decrease of the flow stability in the transonic regime. In this condition, the fluid mode becomes a principal part of the coupling process, which results in the instability of the fluid mode itself or the structural mode, and thus, it is the root cause of different transonic aeroelastic phenomena.

show abstract

Section: Reduced-order Model For the Unsteady Flowmentioning

confidence: 99%

Transonic aeroelasticity: A new perspective from the fluid mode

Gao

Zhang

2020

Progress in Aerospace Sciences

Self Cite

View full text Add to dashboard Cite

show abstract

“…Therefore, they lack robustness with respect to flight parameter variations. To address the ROM adaptation issue discussed previously, the constructed ROM should be parameterized to account for the variations in flight conditions (Kou and Zhang 2017;Winter and Breitsamter 2016;Lindhorst et al 2014;Chen et al 2018;Benner et al 2013). To solve the ROM adaptation issue, the method that interpolates the subspace angles between two POD subspaces was proposed (Lieu and Farhat 2007).…”

Section: Introductionmentioning

confidence: 99%

Active Disturbance Rejection Control for Hypersonic Flutter Suppression Based on Parametric ROM

Chen

Zhao

2020

J. Aerosp. Eng.

View full text Add to dashboard Cite

This paper presents an active disturbance rejection control algorithm (ADRC) to suppress the hypersonic flutter of a three-degreesof-freedom (3-DOF) two-dimensional (2D) airfoil operating in multiparameters space. In this study, two key issues need to be addressed. The first one is the construction of the reduced-order model (ROM) of quasi-steady aerodynamic forces in multiparameter space. To solve this problem, a parametric reduced-order model (PROM) based on the adaptive proper orthogonal decomposition (POD) is developed to efficiently predict the quasi-steady aerodynamic forces at an arbitrary point in the given multiparameters space. The accuracy of the PROM is validated by the direct computational fluid dynamics (CFD) computations for flutter and limit cycle oscillation (LCO) prediction. The second issue is the design of the control law for an aeroservoelastic (ASE) system constructed by a parametric black-box aerodynamic model. Considering the black-box characteristics of a controlled object, we use an error-based feedback control law named ADRC to suppress the flutter instability of the system. Furthermore, in order to improve the rapidity in responses and the robustness of the ASE system, we propose an SVM-ADRC strategy that features embedding the least squares support vector machine (LS-SVM) into the ADRC controller. Hypersonic ASE simulation results show that the proposed SVM-ADRC controller can effectively suppress the hypersonic flutter over a wide range of parameter variations, and is more robust than the standard ADRC controller.

show abstract

“…The experiment results show that deep learning can be applied to unsteady aerodynamic modeling. Kou J. et al [ 30 ] presented a multi-kernel neural network and applied it to a nonlinear unsteady aerodynamic model in varying flow conditions. By comparison with the kernel RBF neural network model, a multi-kernel neural network can be more effectively used in the study of unsteady aerodynamics.…”

Section: Introductionmentioning

confidence: 99%

Modeling of a Novel Coaxial Ducted Fan Aerial Robot Combined with Corner Environment by Using Artificial Neural Network

Chen

et al. 2020

Sensors

View full text Add to dashboard Cite

A novel coaxial ducted fan aerial robot with a manipulator is proposed which can achieve some hover operation tasks in a corner environment, such as switching on and off a wall-attached button on the corner. In order to study the aerodynamic interference between the prototype and the environment when the aerial robot is hovering in the corner environment, a method for the comprehensive modeling of the prototype and corner environment based on the artificial neural network is presented. By using the CFD simulation software, the flow field of the prototype at different positions with the corner effect is analyzed. After determining the input, output and structure of the neural network model, the Adam and gradient descent algorithms are selected as the neural network training algorithms, respectively. In addition, to optimize the initial weights and biases of the neural network model, the genetic algorithm is precisely used. The three-dimensional prediction surfaces generated by the three methods of the neural network, kriging surface and the polynomial fitting are compared. The results show that the neural network has high prediction accuracy, and can be applied to the comprehensive modeling of the prototype and the corner environment.

show abstract

Multi-kernel neural networks for nonlinear unsteady aerodynamic reduced-order modeling

Cited by 54 publications

References 83 publications

Transonic aeroelasticity: A new perspective from the fluid mode

Transonic aeroelasticity: A new perspective from the fluid mode

Active Disturbance Rejection Control for Hypersonic Flutter Suppression Based on Parametric ROM

Modeling of a Novel Coaxial Ducted Fan Aerial Robot Combined with Corner Environment by Using Artificial Neural Network

Contact Info

Product

Resources

About