中文摘要：

民航班机的快速存取记录仪（QAR）记录了大量的飞行和性能参数,QAR数据的约简是对这些数据进行数据挖掘的一个重要环节。针对民航QAR数据的特点及为数据约简后应用数据挖掘算法的需要,在小波变换相关理论分析的基础上,提出了利用小波尺度系数进行QAR时间序列数据约简的方法及其性能分析方法,确定了相应的性能评价指标。实验结果表明了该方法对QAR数据约简的正确性及时域、频域数据特征兼顾,数据的主要特征失真小等优点。

英文摘要：

A great amount of data from aviatic and performance parameters are recorded by the quick access recorder （QAR） equipped on scheduled flights. Reduction of QAR data is an important step through data mining process for these datas. According to the characters of QAR data and for the need of applying data mining arithmetic, based on analysis of relative wavelet transform theory, a method of QAR time sequence data reduction is proposed using wavelet scaling coefficients and corresponding performance analysis method, and corresponding evaluating indicators are defined. Subsequent experiment proves its correctness and superiorities of taking into account of both characters of data in the time domain and freqency domain, less main diagnostic distortion, etc.