Slug flow is one of the most observed gas-liquid flow patterns in pipes. Owing to its high occurrence, the estimation of slug characteristics is essential for many engineering processes. The slug flow characterization is usually carried out by models and correlations previously calibrated with experimental data generated by the interpretation of voltage or instantaneous liquid holdup time-series. Historically, this interpretation required algorithms that depended on subjective parameters, which created high dispersion on the data. This paper proposes a new signal processing analysis, which does not require any subjective parameter. A statistical algorithm is used to calculate the film and slug cut threshold values, the disregard cut value to group slug pulses, and the disregard cut value to remove slug pulses, which are required to determine the slug characteristics. An experimental data set was used to validate the proposed methodology. The consistency check process followed two independent ways, both with good results. Based on the performance against the experimental data, the proposed algorithm is recommended for slug flow characterization.

塞流是管道中最常见的气液流形式之一。由于它的高发生率，对于许多工程过程来说，段塞特性的估计是必不可少的。团状流的表征通常是通过模型和相关性进行的，该模型和相关性以前是通过解释电压或瞬时液体滞留时间序列而生成的实验数据进行校准的。从历史上看，这种解释需要依赖于主观参数的算法，这在数据上造成了高度分散。本文提出了一种新的信号处理分析方法，该方法不需要任何主观参数。使用统计算法来计算胶片和条块切割阈值，忽略条块切割值以分组条块脉冲以及忽略条块切割值以去除条块脉冲，确定弹头特性所需的参数。实验数据集用于验证所提出的方法。一致性检查过程遵循两种独立的方法，均具有良好的结果。基于针对实验数据的性能，建议将所提出的算法用于弹丸流表征。