用于气动导数辨识的试飞数据处理方法研究

飞行试验的测量环境十分复杂,未经处理的试飞数据直接用于飞机的气动导数辨识会降低辨识精度,甚至导致辨识的迭代过程发散或收敛到错误值。提出一套从野值剔除、数据平滑、时延修正到相容性检验的试飞数据处理方法,并将其应用到真实试飞数据的气动导数辨识中,通过对比分析数据处理前后的辨识过程和辨识结果,验证了本文提出的数据处理方法在改善辨识过程收敛特性和提高辨识精度上的有效性。

Flight Test Data Processing Method Research Applied to the Identification of Aerodynamic Derivatives

In flight test, the measuring environment is very complicated. If the flight test data is unprocessed and directly used for the identification of aerodynamic derivatives, it will reduce the accuracy of the identification and even can result in iteration divergence or convergence to the error value. A set of method is put forward, which contains identification and correction of data outliers, smoothing, filtering, time delay correcting, and compatibility checking, the method is applied to the real flight test data of the aerodynamic derivative identifica- tion, through the comparative analysis of identification process and the identification results before and after treatment, to verify the processing method of the flight test data. The method presented is effective at impro- ving the convergence characteristic of the identification process and the accuracy of the identification results.