城市空中交通（Urban Air Mobility，UAM）设想了一种安全高效的航空运输系统，使用高度自动化的飞行器在城市和郊区的低海拔地区运输乘客或货物，旨在缓解日益增长的交通拥堵。为了发展UAM，需要一种具备垂直起降能力、高可靠性、低成本、零排放和低噪声的高速飞行器。电动垂直起降飞行器（electric Vertical Takeoff and Landing, eVTOL）应运而生。作为高度自动化的飞行器，eVTOL需要硬件在环实时仿真在设计和研制环节对其验证。数字模型置信度的提升和飞行控制策略问题是eVTOL开发设计过程中的重要问题。 针对eVTOL的部件进行建模。分别建立了旋翼、螺旋桨、机身、平尾、垂尾和机翼的数字模型。不同构型的eVTOL均可据此搭建。模型验证包含部件和整机两方面。通过将计算结果与试验数据对比，验证了建模的有效性。 针对旋翼或螺旋桨斜向来流时的动态失速，本文建立了翼型动态失速代理模型。经过特征选择和样本选择，动态失速代理模型在测试集上的预测误差小于2.5%。级联跟踪微分器用于从飞行动力学模型提取特征量。代理模型与飞行动力学模型结合，在不同飞行工况下保证了预测精度且实现了硬件在环仿真。 针对螺旋桨/机翼、螺旋桨/螺旋桨气动干扰，结合滑流理论和湍流射流理论，基于速度自相似律假设和试验数据，获得螺旋桨滑流轴向和切向速度分布的解析形式。该模型预测结果相比于Goates方法，轴向速度预测平均误差从16.6%减少到9.6%。切向速度预测的平均误差从28.8%减少到21.1%，与螺旋桨或机翼模型结合可对eVTOL部件间气动干扰进行预测。针对多旋翼可倾转eVTOL，建立了考虑动态失速和气动干扰的飞行动力学实时仿真模型，并给出设计参数，以5 ms步长进行仿真时单步计算约1.75 ms。设计了飞行控制策略，包括高度/速度/航向保持器与短舱倾转策略，并基于振动对人体的影响，评估不同短舱倾转策略。
Urban Air Mobility (UAM) envisions a safe and efficient air transportation system that uses highly automated aircraft to transport passengers or cargo at low altitudes in urban and suburban areas, aiming to alleviate growing traffic congestion. To develop UAM, a high-speed aircraft with vertical takeoff and landing capability, high reliability, low cost, zero emissions, and low noise is needed. The electric Vertical Takeoff and Landing (eVTOL) is developed. As a highly automated vehicle, eVTOL requires hardware-in-the-loop real-time simulation to validate it during design and development. Model fidelity enhancement and flight control strategy are essential in eVTOL development and design. Components of eVTOL are modeled, including the rotor, propeller, fuselage, horizontal tail, vertical fin, and wing. Different configurations of eVTOL can be built based on these components. The Validation of modeling is verified by comparing the simulation results of the parts and the whole aircraft with the experimental data. Dynamic stall occurs when the rotor or propeller works in the oblique flow. This paper establishes a dynamic stall surrogate model. After feature selection and sample selection, the prediction error of the dynamic stall surrogate model on the test set is less than 2.5%. The cascaded tracking differentiator extracts the features from the flight dynamics model. Combining the surrogate model and flight dynamics model ensures prediction accuracy under different flight conditions and achieves hardware-in-the-loop simulation. For propeller/wing and propeller/propeller aerodynamic interferences, the mathematical expressions of propeller slipstream axial velocity and tangential velocity distributions are obtained based on the assumption of velocity self-similarity law, experimental data by combining slipstream theory, and turbulent jet theory. Compared with the Goates method, the model reduces the average error of axial velocity from 16.6% to 9.6%, while the average error of tangential velocity is reduced from 28.8% to 21.1%. The combination with the propeller or wing model can predict the aerodynamic interferences between eVTOL components. For the multi-rotor tiltable eVTOL, a flight dynamics real-time simulation model considering dynamic stall and aerodynamic interference is established, and the design parameters are given. The computation time required for the model simulation step of 5 ms is about 1.75 ms. The flight control strategy is designed, including altitude/speed/yaw holder and nacelle tilt strategy. Different nacelle tilt strategies are evaluated based on the effect of vibration on the human body.
