基于EMD的稠油油井出砂信号特征识别

在进行基于振动的稠油油井出砂监测试验中,为更好地识别出砂信号,提高信噪比,准确建立出砂参数与出砂信号间的关系,将经验模态分解(EMD)应用到出砂信号的特征提取中。采用EMD方法将出砂信号分解成固有模态分量,并通过峭度分析和能量分析,在这些分量中得到包含出砂振动信号特征的3个IMF分量。通过对IMF分量进行频域分析,得到出砂信号频域范围。通过室内出砂监测试验,验证了EMD方法对出砂信号特征识别的有效性,为现场出砂监测信号处理提供了一种新的思路。

EMD-based Signal Feature Identification of Sand Production in Heavy Oil Well

To better identify sand production signals,improve signal-to-noise ratio(SNR),and precisely establish the relation between sand production parameters and signals,the empirical mode decomposition(EMD) was applied to feature extraction of sand production signals in the vibration-based monitoring test of heavy oil well sand production.The EMD method was adopted to decompose sand production signals into intrinsic mode function.Three IMF components containing sand production vibration signal feature were derived from the components through kurtosis analysis and energy analysis.The frequency-domain analysis of the IMF component was conducted to obtain the frequency-domain scope of sand production signals.The laboratory monitoring test of sand production verified the effectiveness of the EMD method in identification of sand production signal feature,providing a new orientation for the processing of field sand production monitoring signals.