一种飞行数据预测方法

Prediction Method for Flight Data Processing

为了解决实际飞行试验中机载系统与GPS采样率频率相异,难以对飞行状态实时估计和性能导航进行计算等问题,通过对GPS进行预测来对数据统一化处理。如果机载系统采样频率是GPS的整数倍且GPS输出时间对应于机载系统采样时刻,给出了一种基于模型的预测方法。对于一般情况,采用了时间序列的混沌多项式预测模型,对实测数据进行相空间重构的基础上,选取最优邻近点进行预测。实际飞行试验数据分析结果表明,给出的预测方法是有效的。

It is hard to estimate flight states and realize navigation directly since the global positioning system （GPS） sample frequency is different from that of the inertial navigation system （INS） in practical flight test. To solve the problem,GPS data prediction was used to uniformly process data. In the new scheme,a model based prediction method is proposed when the sample frequency of INS is integer multiple to GPS and GPS output time is corresponding to the sample time of INS. Otherwise,both the chaotic polynomial and the optimal proximal point prediction method of time series are used for the data reconstruction by means of phase space. Flight data processing shows that the prediction method is effective.