对于周期脉冲序列的信号分选识别，可采用序列搜索的时域方法[1，2]。本文将大周期内完全重复的脉冲序列分解为多个“单周期信号”．并把它们均定义为周期性信号．同时，提出了序列搜索的一种软件实现方法及密集视频脉冲信号分选的变步长、信步长快速搜索法．在此基础上，本文针对序列搜索分选法的可信度计算公式进行了推导和分析，证实了这种方法的可行性和有效性． For periodic pulse sequence signal classification and recognition, time-domain search method [1, 2] can be used. In this paper, a completely repeated pulse sequence in a large period is decomposed into multiple “single-period signals.” And define them both as periodic signals. At the same time, a software implementation method of sequence search and variable step length and fast searching method of signal strobe are proposed. On this basis, this paper deduces and analyzes the formula for calculating the credibility of sequential search and sorting method, and proves the feasibility and effectiveness of this method.