大数据驱动的飞机翼身对接装配位姿分析方法

Data-driven Method of the Wing Position Analysis for the Aircraft Assembly

针对大飞机翼身装配位姿分析问题,提出了基于调姿定位器轴力数据的翼身对接位姿分析方法。首先针对翼身对接位姿分析中的各阶段数据进行数据建模,构建包含元元模型、元模型、元数据和用户数据的四层结构的数据模型。在此基础上,设计面向数据高噪声的定位器数据处理方法,实现数据清洗,并基于数据立方体设计数据转换方法,实现定位器数据从面向采集与存储到面向形变分析的转换。最后,基于Ko位移理论,设计机翼与中机身位姿的分析模型,对某机型的左翼安装角、上反角和后掠角进行误差分析。相比于传统方法,数据驱动的方法能够实时获取定位器数据,实现时序数据分析,可大幅降低误差测量与分析周期,提升整机装配过程中的误差变化透明程度。

In order to analysis the position change of wing during wing-fuselage joining assembly,this paper introduces a method based on the data collected form three-axis positioners. First,to consider the each life stage of its assembly,we built a four layered data model system including meta-meta model layer,meta model layer,meta data layer and user data layer. Then,on base of this model,the specific method was used to clean the data. The data cube transformation method was used to transform the collection oriented data into analysis oriented data. In the wing-fuselage assembly,the analysis method based on Ko displacement theory was used to analysis the deviation of wing＇s setting angle,positive-dihedral angle and sweep angle. Compared with other method,this data-driven method used the real time series data collected from positioners to analysis and monitor the position change of wind. It reduced the analysis time and improved the transparency of assembly deviation.