针对不等长序列数据的关联问题,提出基于滑动窗口的最优匹配增权法不等长序列相似度度量算法.以较短序列作为滑动窗口遍历较长序列得到一组滑动相似度,利用这组相似度形成最优权重,加权得到不等长序列的相似度,并根据相似度大小对序列数据进行关联判决,以解决截断法相似度度量仅能反映截断序列局部相似度的问题.仿真实验验证了所提出算法对不等长序列数据关联的有效性,并对序列长度和量测误差等因素对相似度度量和关联效果的影响进行了讨论.
An optimal matching increasing weight algorithm for the unequal length sequence similarity measurement based on the sliding window is proposed to solve the unequal length sequence data association, which uses shorter sequences slide longer ones to get slidable similarity, forming the optimal weight with this similarity at the same time, then weighting the slidable similarity to get the unequal length sequence similarity. According to the degree of the sequence similarity, the judgment of the association of unequal length sequence data is obtained, which solves the local similarity problem of the truncated measurement algorithm. Simulation experiments show that the proposed algorithm can associate unequal length data effectively and also discuss the influence of the variation of sequence and measurement error on the sequence similarity and the association effect.