摘要： 针对多元时间序列搜索的实际特点，提出WSTB方法，以线性分段建立时间序列曲线箱。创立具有相似形状的时序子序列箱，建立相应的索引，对给定序列和相似序列距离的快速计算，根据查询的时间序列的特征确定相应的权重，不需要逐个检查子序列箱内容就可以进行快速索引。采用实际飞行数据进行适应性验证，验证了该算法的通用性和有效性。

关键词: 数据挖掘, 线性分段, 相似性, 飞行数据

Abstract: A WSTB-based algorithm for similarity search is proposed which is based on the piecewise linear representation to solve the high-dimensional time series searching. The subsequence bin for time series is built and then the index of the bin is built. The distance of the given sequence and similar sequence is calculated. The weighted coefficient for every sequence is decided on the character. So the inquiry can be implemented without checking the content of the bin. The quantity of the WSTB calculation which is gotten from compare one by one is avoided. The searching efficiency can be improved obviously. The currency and efficiency of the algorithm is proved with high-dimensional flight data in this paper.

Key words: Data mining, Piecewise linear representation, Similarity, Fight data