Icing encountered by an aircraft in flight can pose a great threat to flight safety, which is the foremost concern in aviation. Ice accretion has nonlinear and coupling properties, and therefore, a conventional envelope protection system cannot successfully deal with the condition. A safe flight envelope based on reachability analysis is proposed in this paper as a basis for designing an online envelope protection system for aircraft under icing conditions. This system uses a neural network method to identify changes in the aerodynamic coefficients and to classify the degrees of icing and uses a database-driven approach to solve the onboard safe flight envelope computation problem. The database of icing safe flight envelopes is created offline and can be retrieved in real time. The Research Civil Aircraft Model (RCAM) encountering icing conditions in the climbing phase was taken as an example to verify the feasibility of the online safe flight envelope protection system. The simulation results showed that the system can classify the degrees of icing efficiently, prevent the aircraft from deviating from the safe flight envelope, guide it back to the envelope along the fastest path, and reduce the risk of loss of control under icing conditions.

中文翻译：

---

基于可达性分析的结冰飞机在线安全飞行包线保护

飞机在飞行中遇到结冰会对飞行安全造成极大威胁，这是航空界最关心的问题。积冰具有非线性和耦合特性，因此传统的包层保护系统无法成功应对这种情况。本文提出了一种基于可达性分析的安全飞行包线，作为设计结冰条件下飞机在线包线保护系统的基础。该系统使用神经网络方法来识别空气动力系数的变化并对结冰程度进行分类，并使用数据库驱动的方法来解决机载安全飞行包线计算问题。结冰安全飞行包线数据库离线创建，可实时检索。以爬升阶段遇到结冰条件的研究民用飞机模型（RCAM）为例，验证了在线安全飞行包线保护系统的可行性。仿真结果表明，该系统能够有效地对结冰程度进行分类，防止飞机偏离安全飞行包线，引导飞机沿最快路径返回包线，降低结冰情况下失控的风险。