基于多源数据融合的飞行风险时空分析与预警研究

Flight risks characterized by the Quick Access Recorder (QAR) data are closely related to meteorological and terrain factors. There, however are few studies on fusing all these data sets due to the difficulties in the QAR data accesses. The Flight Quality Control Base Station of China Civil Aviation has been constructed to collect all the QAR data in China, which provides a good data foundation for this project. In this project, we will carry out studies on spatio-temporal analysis and early warning of flight risks with fusing multi-source data sets. All in all, Data fusions will be fundamental, spatio-temporal analysis be the means and early warning of flight risks be the goal. In details, the contents of this project include: 1) Fusing multi-source data sets with QAR data; 2) Exploring spatio-temporal patterns of flight risks; 3) Analyzing spatio-temporal correlation factors and modelling relationships with flight risks; 4) Accurate early warning and sSpatio-temporal visualization and early warning of flight risks. This project will fully utilize the QAR data to explore approaches and models for accurate early warning and spatio-temporal visualization of flight risks, specifically by introducing muti-scale spatio-temporal analytic techniques. The outcomes from this project will provide technique supporting and decision making services for flight crew preparation, flight dispatching, air traffic control commanding and dispatching, new airport allocating and flight route optimization, which will make propose effective methodologies in fusing and analyzing multi-source spatio-temporal data sets with QAR data to reduce and avoid flight risks, and it will make a fundamental contribution to enhancing the flight safety level of China civil aviation.

飞机快速存取记录器QAR数据所表征的飞行风险与气象、地形等要素密切相关。由于学界难以获取行业QAR数据，有关QAR数据与气象、地形等数据进行融合与时空分析的研究很少。中国民航飞行品质监控基站的建设首次汇聚了国内全行业的QAR数据，为本项目奠定了坚实的数据基础。本项目拟融合气象、地形数据，以数据融合为基础、以时空分析为手段、以风险预警为目标，开展飞行风险时空分析与预警研究。具体内容包括：1）多源异构QAR大数据融合技术研究；2）飞行风险挖掘分析与时空分布模式探索；3）飞行风险时空关联要素分析与关系建模；4）飞行风险精准预警与时空可视化。本项目将有效利用全行业QAR数据，引入多尺度时空异构分析技术，探索精准的飞行风险预警模式和时空立体可视化，为飞行机组准备、签派放行、空管指挥调度、新机场选址和航线优化等提供技术支持和决策服务，从而为减少规避飞行风险进一步提升我国民航飞行安全水平做出基础性贡献。

飞机快速存取记录器QAR数据所表征的飞行风险与气象、地形等要素密切相关。由于学界难以获取行业QAR数据，有关QAR数据与气象、地形等数据进行融合与时空分析的研究很少。中国民航飞行品质监控基站的建设首次汇聚了国内全行业运输飞机的QAR数据，为本项目奠定了坚实的数据基础。本项目融合气象、地形数据，以数据融合为基础、以时空分析为手段、以风险预警为目标，开展飞行风险时空分析与预警研究。具体内容包括：1）多源异构QAR大数据融合技术研究；2）飞行风险挖掘分析与时空分布模式探索；3）飞行风险时空关联要素分析与关系建模；4）飞行风险精准预警与时空可视化。本项目取得的重要研究进展概括为两个成果，一是出版了专著《飞行安全时空大数据理论与实践》，二是研发了预警平台及APP。本项目有效利用全行业QAR数据，引入多尺度时空异构分析技术，探索精准的飞行风险预警模式和时空立体可视化，为飞行机组准备、签派放行、新机场选址和航线优化等提供技术支持和决策服务，从而为减少规避飞行风险进一步提升我国民航飞行安全水平做出基础性贡献。