| 邸亚洲1，秦永元1，尚希良2，曲建岭2 |
（1. 西北工业大学 自动化学院，陕西 西安710072；2. 海军航空工程学院 青岛分院，山东 青岛266041） |
 摘要：提出利用多项式回归算法对飞参记录数据存在的随机量测误差、野点以及数据丢失等现象进行有效的数据预处理，算法在消除量测误差、剔除和补正野点、补充丢失的数据及数据平滑等方面均具有较高的精度和可靠性并已有效地应用在多型飞机飞参记录数据预处理工作中。利用该算法在Matlab环境下对飞参记录的航姿系统俯仰通道部分数据预处理过程进行了仿真。 |
| 关键词：多项式回归；飞行参数记录系统；数据预处理；野点 |
| 中图分类号：TP274 文献标识码：A 文章编号：1000-8829(2008)04-0021-02 |
| Research on Data Pre-Processing of Flight Data Recorder Based on Polynomial RegressionSystem |
| DI Ya-zhou1， QIN Yong-yuan1， SHANG Xi-liang2， QU Jian-ling2 |
| (1. Department of Automatic Control,Northwestern Polytechnical University,Xi’an 710072,China;2. Naval Aeronautical Engineering Academy Qingdao Branch,Qingdao 266041,China) |
| Abstract：The flight data which has been recorded in flight data recorder systems(FDRS) is always contaminated by measurement noise,singular values and missing data,which has to be pre-processed.For this pre-processing,an algorithm of polynomial regression which can reduce the measurement noise,correct the singular values and recover the missing data efficaciously is proposed.Meanwhile this algorithm has higher precision and dependability.This method has been applied in the field of data ground pre-processing of FDRS which has been installed in multi-type airplane.This algorithm is used to fitting the data of airplane pitching angle by Matlab program. |
| Key words：polynomial regression;flight data recorder systems;data pre-processing;singular values |
|
|
