跨声速颤振是一种经典的气动弹性问题，可能会导致飞机结构损坏，对新 型号的研发进程产生影响。随着计算流体力学的发展，数值仿真已成为研究跨 声速颤振及其抑制方法的重要手段。 本文采用基于非定常雷诺平均 N-S 方程的流固耦合时域模拟方法，研究了 弹性翼型在跨声速状态下的颤振问题，主要进行了以下工作： （1）颤振降阶模型的建立：直接 CFD/CSD 耦合计算需要消耗大量的时间， 还不利于颤振机理分析。本文通过结合 ARX 线性模型与 BP 神经网络非线性模 型建立了非定常气动力的降阶模型，极大降低了颤振的计算成本；对基于 ARX 线性模型气动弹性状态空间方程开展特征根分析，辨识了颤振耦合机理。基于 ARX 结合 BP 神经网络建立的非线性降阶模型，实现了对大攻角颤振的预测 （2）提出了一种跨声速颤振的控制方法：由于颤振现象的危害，如何对颤 振进行控制也是一个重要的课题。蒙皮振动技术是利用蒙皮局部结构的主动变 形，对翼型局部流场结构产生影响，进而提升飞行器相关性能。本文采用蒙皮 振动的方式，对颤振达到了良好的抑制效果。 （3）发展了一种二自由度能量图的颤振分析方法：颤振的本质是结构吸收 了空气中的能量，当能量输入为正时，结构振幅就会越来越大，能量输入为负 时，结构振幅就会越来越小，当能量输入为零时，结构变形就会进入等幅振荡， 因此，通过计算各个工况下的能量，就可以得知颤振的发展方向。本文在现有 的单自由度能量图法的基础上，发展了二自由度能量图法，使之能够适用于跨 声速二自由度颤振的分析

Transonic flutter is a classic aeroelastic problem, which may lead to the destruction of aircraft structure and often affect the development process of new aircraft. With the development of computational fluid dynamics, numerical simulation has become an important means of transonic flutter research and its control. In this paper, the unsteady Reynolds mean N-S equation (URANS) is used to study the transonic flutter of elastic airfoil through the fluid-structure coupling time domain simulation method. The main work is as follows: (1) Establishment of flutter reduced order model: CFD/CSD direct coupling calculation needs a lot of time, which is not conducive to mechanism analysis . In this paper,two reduced order models of unsteady aerodynamic forces are established based on ARX linear model and BP neural network nonlinear model, which greatly reduces the calculation cost of flutter; The characteristic root analysis of the aeroelastic state space equations based on the ARX linear model is performed to identify the flutter coupling characteristics. (2) A control method of transonic flutter is proposed. Because of the harm of flutter phenomenon, how to control flutter is also an important topic. The active vibration technology of skin is to use the active deformation of local skin structure to influence the local flow field structure of airfoil, and then improve the relevant performance of aircraft. In this paper, the active vibration of skin is adopted, and the flutter is well suppressed. (3) A flutter analysis method of two-degree-of-freedom energy diagram is developed . The essence of flutter is to absorb the energy in the air. When the energy input is positive, the amplitude will become larger and larger. When the energy input is negative, the amplitude will become smaller and smaller. When the energy input is zero, it will enter into constant amplitude oscillation. Therefore, the development direction of flutter can be known by calculating the energy under various working conditions. Based on the existing single-degree-of-freedom energy diagram method, the twodegree-of-freedom energy diagram method is proposed in this paper to make it suitable for the analysis of transonic two-degree-of-freedom flutter.