应用模糊熵和相似度的管道泄漏检测与定位
Applying Fuzzy Entropy and Similarity Dgree to Detecting and Locating Pipeline Leakage

动态压力波信号的异常识别是声波法输油管道泄漏检测和泄漏定位的关键。从时间序列角度出发,提出一种基于模糊集理论的输油管道泄漏检测定位方法。该方法首先采用小波分解和半硬软阈值法去除动态压力波信号噪声;其次选取固定长度的正常动态压力波信号子序列作为模板序列,并对现场采集所得动态压力波信号选取相同长度的子序列作为被匹配序列,构造两子序列信号所对应功率谱的模糊集并计算其相似度;最后根据模糊相似度判定输油管道是否存在泄漏现象,若发生泄漏,可利用检测信号和模板信号的模糊熵差确定泄漏位置。实验表明,建议方法简单、有效且定位精度高。
The recognition of abnormal dynamic pressure wave signal is key to pipeline leakage detection and location with acoustic wave. Starting from time sequence, we propose a petroleum pipeline leakage detection and location method based on the fuzzy set theory. Firstly, an adaptive filter is designed to remove the noise of a dynamic pressure wave signal with wavelet decomposition and semi-hard and soft threshold. Secondly, the fixed-length normal dynamic pressure wave signal sequences are selected as template sequences, and the dynamic pressure wave signals with the same-length sub-sequence collected on site are selected as matching sequences; then the fuzzy sets of two sub-sequence signals' power spectra are constructed to calculate their fuzzy similarity degrees. Finally, we use the fuzzy similarity degrees to judge if petroleum pipeline leakage occurs; if yes, we locate the leakage by using the fuzzy entropy difference between normal and collected dynamic pressure wave signal sequences of the petroleum pipeline. The experimental results show that our method is simple, effective and has high location accuracy