飞机机动划分的数据挖掘方法

An Aircraft Maneuver Partition Method Based on Data Mining

飞机机动划分是将飞行数据分解成若干具有明确物理意义的机动动作子序列的重要前处理方法,也是健康监控、飞行模拟、飞行品质评估等研究工作的必要步骤。结合数据挖掘技术提出一种自动的飞机机动划分方法,该方法根据法向过载数据的趋势提取出飞行数据中的机动片段,并用ISODATA聚类将机动片段归并为若干分类,可以证明每个分类代表一种机动动作。将该方法分别应用于小规模飞行数据与大规模飞行数据中能够识别并正确划分至少89%的机动动作,证明该方法有效且满足工程精度要求。

that are m An aircraft maneuver partition method is to divide flight data into several maneuver action sub-sequences eaningful in physics, being essential for health monitoring, flight simulation and flight quality evaluation. Based on data mining, we propose an automatic maneuver partition method, which extracts the maneuver segments of flight data according to the trend of normal overload data and then uses the iterative self-organized data analysis algorithm （ISODATA） to cluster the maneuver segments into some classes. We prove that each class represents a maneuver action. The maneuver partition method is applied to small scale flight data and large scale flight data re- spectively and can recognize and correctly partition at least 89% of maneuver actions, indicating that the method is effective and satisfies the requirements for engineering accuracy.