基于电成像测井图像识别求取地层产状的新方法

正确识别地层界面并准确计算地层产状对于油气勘探非常重要。针对国内测井领域地层界面识别及产状求取方法存在的分辨率低、精度低及速度慢等问题,提出一种基于电成像测井图像识别求取地层产状的新方法:应用曲线斜率拟合法提取地层界面对应的正弦曲线,以提高地层产状计算的速度;采用双窗口检测、交互移动技术以及优选控制准则来提高地层产状计算的精度;通过设计不同步长、窗长组合分别检测地层宏观构造面与微观沉积层理面。理论地层模型与实测资料应用结果表明,该方法求取的地层产状较好地反映了地层实际情况。

Accurate identification of the formation interface and extraction of the formation occurrence for hydrocarbon exploration are very important.At present,there are some problems such as low resolution,low accuracy and low efficiency in the formation interface detection and its dip extraction.Therefore,a new method to detect formation occurrence from image recognition of electric imaging logging is proposed in this paper.The sine curves of the formation interfaces are fitted with the curve fitting method to improve the calculation efficiency of the detected formation interfaces.The double window monitoring,interactive mobile technology and the prefered criterions are used to improve the accuracy of bed occurrence.Both macro-structured interface and micro-sedimentary bedding interface are respectively extracted by designing the combination of different step sizes and window sizes.The results of the theoretical formation models and the measured data detected with the method give various types of the formation interfaces and formation occurrence,which accord with those detected by the foreign advanced software.

国家自然科学基金项目(41374148)资助。