中文摘要：

研究航天员呼吸强度时序模式相似性比较方法。利用小波变换将数据样本中的信号和噪声分离,将样本信号划分成多个连续的单调区间,用年n阶多项式对区间进行分段拟合,由拟合多项式的各阶系数、区间宽度、信噪比组成该分段区间的特征向量,由此得到所有数据样本的特征向量序列。小波分析算法可以有效地分离噪声信号,多项式函数能够很好地拟合数据样本,用特征向量的余弦距离可以对独立或连续区间的相似度进行准确地比较。根据上述方法,实现了时间序列的相似性搜索,并用于航天员呼吸强度时序变化模式的识别。

英文摘要：

To study the approach for Similarity Comparison of sequential pattern of Astronaut＇ s Respiratory Intensity. The signal and the noise separated by wavelet analysis, the whole data sample was divided into many continuous intervals in which the signal had monotone variation trend. Each interval was fitted by n-degree polynomial and its＇ eigenvector was made up of n ＋ 1 coefficients of the n-degree polynomial, width and SNR. All the eigenvectors composed a series to express the whole sample. Based on wavelet algorithm, the noise is filtered effectively, the polynomial fitted the sample to an adequate precision, and the cosine distance of two intervals＇ eigenvector exactly calculated the similarity of them. According to such representation, this paper can implement the similarity searching of time series and sequential pattern recognition of Astronaut＇s Respiratory Intensity.