激波-激波干扰问题是吸气式高超声速飞行器发动机的内外流耦合流动中最 具挑战性的难题之一。由激波相互作用所产生的激波、膨胀波、剪切层和射流等 复杂流动结构，一方面会在干扰区域附近与壁面边界层相互作用，给物体表面带 来极高的热载荷和力载荷，给飞行器的飞行安全和使用寿命都带来严峻的考验； 另一方面，激波-激波干扰也可能会伴随一系列的层流-湍流转捩、化学非平衡过 程，非定常振荡等复杂现象，使得问题的难度进一步提升。其中，激波-激波干扰 的流场预测是研究者们重点关注的科学问题之一，具有重要的科研意义和工程应 用价值。本文主要通过数值模拟和机器学习算法，建立了具有显式解析表达式的 激波干扰流场预测模型以及激波干扰类型转变准则。该准则能够正确预测壁面压 力和热流的分布规律。另外，本文通过理论方法和数值模拟，分析了不同气体模 型下的激波干扰流场结构和流场参数，揭示了高温真实气体效应对激波干扰流场 构型和流场参数的普适影响规律。取得的主要创新性研究性成果如下： (1) 针对圆柱诱导的弓形激波和入射斜激波的干扰问题，分别基于量热完全 气体模型和仅考虑振动激发的热完全气体模型，数值求解二维可压缩 NS 方程， 得到一系列二维激波干扰流场的数值模拟结果，形成斜激波-弓形激波干扰流场 数据库，为机器学习过程提供充足的数据样本。 (2) 基于 MBB 算法建立了不同气体模型下的 IV 型激波干扰流场结构的预 测模型。该模型具有显式的解析表达式，能给出 IV 型激波干扰流场的特征位置 参数（上下三波点位置）的坐标公式，从而进行超声速射流结构的几何结构的预 测。数值结果与机器学习的预测结果表明，该预测模型具有较高的精度。 (3) 基于 MBB 算法，建立了不同激波干扰模型下的激波干扰类型转变准则 分界面的三维曲面方程。同时，利用方程在特定平面上的投影，得到了随来流参 数变化的不同气体模型下的干扰类型转变准则曲线，清晰刻画了来流参数对干扰 类型转变过程的影响规律并开展了大量的数值模拟，验证转变准则的性能。对该 准则上的临界工况的激波干扰结构和类型以及壁面压力和热流的分布规律进行 预测，结果表明本文所提出的准则具有很高的准确性和可推广性，可以用来快速 预测工程计算中所发生的激波干扰类型，对采用不同气体模型计算时所带来的压 力和热流分布的差异进行合理的预测。 (4) 通过大量的数值模拟结果分析，揭示了高温气体效应对典型的激波干扰 流场参数和流场构型的影响机理。当改变来流条件时，高温气体效应对弓形激波 波后的压力和马赫数的影响并不显著，振动激发使得气体分子的部分平动能转化 为振动能，故导致波后温度更低。此外，由于振动激发，气体比热比减小，可压 缩性增强，故激波脱体距离减小，整体射流结构更靠近圆柱壁面。

Shock-shock interference prediction is one of the most challenging problems in hypersonics. There are a lot of complex flow structures such as shock waves, expansion waves, shear layers and jets generated by the shock-shock interference. On the one hand, will interact with the wall boundary layer near the interference area, bringing extremely high thermal loads and forces to the surface of the object. On the one hand, the shock wave and boundary layer interaction will cause the separation and reattachment of the boundary layer in the vicinity of the interference area, which will bring extremely high thermal loads to the surface of the aircraft. Furthermore, it brings huge challenges to the flight safety and service life of the aircraft. On the other hand, shock-shock interference may also be accompanied by a series of complex phenomena such as laminar-turbulent transitions, chemical non-equilibrium processes, and unsteady oscillations, which further increases the difficulty of the problem. The flow field prediction problem based on shock-shock interference has important scientific research significance and engineering application value. Therefore, through numerical simulation and machine learning algorithm, this paper successfully established the shock-shock interference flow field prediction model and the transition criterion of different shock-shock interference types, and proved that the criterion can reasonably predict the interference law of wall pressure and heat flux. In addition, through theoretical methods and numerical simulations, this paper analyzes the flow field structure of shock-shock interference and flow field parameters under different gas models, and reveals the universal influence of the high-temperature real gas effect on the shock-shock interference flow field and flow field parameters. The main innovative research results obtained are as follows. (1) In this paper, we analyze the effect of high-temperature gas effects on the structure of the shock interference and the flow field parameters, especially of the type IV shock interference, based on the calorimetric complete gas model and the thermal complete gas model, respectively, by numerically solving the viscous two-dimensional compressible Navier - Stokes equations for the cylindrical-induced bow shock and oblique shock interference problems. Multiple sets of numerical simulation results provide sufficient data samples for the machine learning process. (2) Based on the MBB algorithm, the prediction model of the flow field structure of type IV shock-shock interference under different gas models was successfully established. In this paper, the formulas of the characteristic position parameters (the positions of the upper and lower triple points) of the type IV shock-shock interference flow field are given, and the prediction model of the geometric structure of the supersonic jet structure is shown in this paper. And compared the numerical results with the results of machine learning, it is showed that the prediction model with high accuracy reflects the superiority of the algorithm in fluid mechanics data processing. IV (3) Based on the MBB algorithm, the transition criteria of shock interference types under different shock interference models are proposed, and the three-dimensional surface equations of the type transition criteria are given, and the four sets of equations are projected on a specific plane, and the variation of the parameters with the incoming flow is obtained. The interference type transition criterion curves under different gas models, which describe the influence law of the incoming flow parameters on the interference type transition process. (4) The transition criterion proposed above is verified, and the structure and type of shock-shock interference and the distribution of wall pressure and heat flux in the critical condition on the criterion are predicted, which proves that the criterion proposed in this paper has certain accuracy and reliability. Generalizable, it could be used to quickly predict the type of shock-shock interference that occurs in engineering calculations, and make reasonable assumptions about the differences in pressure and heat flux distribution caused by different gas models. (5) The paper reveals the influence mechanism of the high temperature gas effect on the typical the flow field parameters of shock-shock interference and flow field structure. When the incoming flow conditions are changed, the effect of high temperature gas on the pressure and Mach number behind the bow shock wave is not significant. However, the vibrational excitation converts part of the translational energy of the gas molecules into vibrational energy, resulting in a lower temperature of the flow field. In addition, due to the vibration excitation, the γ of the gas is reduced, and the compressibility is enhanced, so the distance of the shock wave is reduced, and the overall jet is closer to the cylindrical wall.