Research On Fault Detection Methods Of Seismic Attribute Analysis

The goals of searching for oil and gas have become more and more refined with the indepth development of petroleum exploration,and the distribution and morphology of faults play an increasingly important role in the identification and description of oil and gas reservoirs.Through the efforts of the past few decades,people have been able to identify faults with large distances and drops with the help of three-dimensional seismic exploration technology.Especially after the application of coherence cube technology,the accuracy of fault description has been greatly improved.Nevertheless,they are still far away from the goal of geologists.The identification and description of small faults has become a difficult point in seismic data processing and interpretation.For the detection and recognition of small faults,the traditional technology is not effective.Therefore,the systematic study of faults,especially the detection and recognition technology of small faults,has very important theoretical significance and practical value.Through seismic attribute analysis,the three methods complement each other to finely describe the characteristics of the fault.Finally,the information fusion method is used to improve the accuracy of detecting faults based on seismic attributes.Based on the research progress of seismic attributes and fault recognition technology,this paper studies the detection of faults through gradient structure tensor attributes,similar attributes and local structure entropy attributes: the gradient structure tensor attributes are compared with conventional coherence methods through models and actual data.,The gradient structure tensor method has good stability,is less affected by the structural background,and contains richer discontinuity information to extract small fractures;similar attributes mainly introduces the algorithm principles of traditional coherence algorithms,and proposes a method based on The direction-oriented similarity algorithm analyzes the connection and difference between the first-generation seismic coherence cube,the second-generation seismic coherence cube,and the third-generation seismic coherence cube algorithm,as well as their respective adaptation conditions.After adding the directional weight factor,it is suppressed to a certain extent.The influence of the stratum is used to extract the main fault;the local structural entropy attribute can effectively suppress the noise through the detection of the model and the actual data;then through the fault attribute fusion,the three methods complement each other to obtain richer and clearer fault information.The results of fault distribution have consistent trends and reveal the characteristics of fault distribution more clearly,with obvious application effects.In the full text,close to the three-dimensional post-stack seismic data of a certain area in western my country,through the data processing of three methods,complement each other,and finally realize the detection of faults in this area,which verifies the effectiveness of this method.