复杂环境下基于机器学习的空管指挥安全协同监控理论与方法研究

空管安全是国家空域安全的基石，管制指挥是空管安全的关键。空管环境日益复杂，给空管安全带来极大隐患，迫切需要开展复杂环境下新的空管指挥安全协同监控理论与方法的研究。本项目旨在建立AI空管安全监督员模型，突破人-机协同交互模型、空管指挥知识表示难题，解决人机协同的空管指令实时安全监控、飞行冲突探测与解脱等问题。研究内容：首先，研究空管指挥人机协同安全监控模型，奠定项目的理论基础；其次，研究空管指挥知识表示与学习方法、空管指令安全协同监控、空管飞行冲突探测与协同调配关键技术；最后，研发原型系统并开展实例评估，验证提出的理论和关键技术。通过提供具有自主知识产权的理论技术，本项目尝试从科学角度回答能否利用机器学习有效解决复杂环境下的空管指挥安全协同监控、人机可信赖交互以及可解释问题，具有重要的理论意义;对于提高空管指挥安全监控能力与智能化水平，提升空管装备国产化水平具有十分重大的应用价值。

Air Traffic Control is the key to aviation safety, and even national airspace security. The ATC environment is becoming more and more complex, which brings great hidden dangers to air traffic control safety. It is urgent to research new air traffic control human cooperative monitoring theory and methods in complex environments. This project aims to establish an AI ATC safety supervisor model, break through the human-machine collaborative interaction model, air traffic command knowledge representation problems, and solve human-machine coordinated air traffic control real-time safety monitoring, flight conflict detection, and resolve the problem.. Research contents: Firstly, study the human-machine cooperative safety monitoring model of air traffic command to lay the theoretical foundation of the project; Second, study the knowledge representation and learning methods of air traffic command from big data, air traffic command safety monitoring, air traffic flight conflict detection, and coordination resolution technology; Finally, the prototype system is developed and evaluated by examples to verify the theory and key technologies proposed. . By providing theoretical technologies with independent intellectual property rights, this project attempts to answer from a scientific perspective whether it is possible to use machine learning to effectively solve air traffic command security collaborative monitoring, man-machine reliable interaction, and interpretable problems in complex environments, which has important theoretical significance. It is of great application value to improve the air traffic control command safety monitoring capability and intelligence level and to enhance the localization level of air traffic control equipment.