Oil-gas Reservior Prediction Of Multi-wave Seismic Based On Fuzzy Cluster Fusion Analysis

Seismic data conceals abundant stratigraphic information, include structural feature, lithologic variation, reservoir physical properties and geophysical features.Every attribute is a subset of seismic dat, so using seismic attribute analysis technology to carry on the reservoir prediction has become one of the most widely used and effective means. Anisotropic and heterogeneous underground medium results in complex seismic geological features, which determines that comprehensive logging, seismic geological data should be used in the practical application. Comprehensive prediction of reservoir by multiple perspectives and many means can achieve advantageous complementarities,and improve the robustness and effectiveness of the attribute analysis technology applied into reservoir prediction.Based on multi-attribute analysis, this paper adopts multi-wave and multi-attributes analysis technology and add to converted shear wave data,which can provide more abundant seismic wave field information and motion information. Besides, Converted shear wave data has its unique advantages in the aspects of fracture detection, gas bearing prediction and fluid properties evaluation. The combination of P-wave and converted shear wave data for multi-attribute analysis is bound to be better than conventional P-wave attributes analysis. multi-wave and multi-attributes analysis can better identify whether the "bright spot" is true or false, and reduce the multiple solutions of data interpretation, and further improve the accuracy of reservoir prediction and fracture detection.First, 18 seismic attributes are extracted along the layer ,based on the geologic-seismic features. Taking the redundancy and waste of information among multiple seismic attributes, a sensitive and relatively independent attribute subset is optimized for the response of the oil and gas reservoirs by means of cluster analysis and expert experience. Second, multi-wave compound attributes and the sweetness attributes are constructed, which has obtained better results in practical application. And Fractures prediction of multi-wave seismic attributes has enriched the method of oil and gas reservoir prediction from another point of view, thus improving the accuracy of reservoir prediction and fractures prediction.Finally, this paper focuses on the basic principle of fuzzy C mean clustering algorithm and the steps of fusion, and realizes the multi-attribute fusion based on FCM algorithm through computer programming, establishes the nonlinear relationship between fusion attributes and reservoir characteristics. Based on the above analysis, finally realizes the oil-gas accumulation prediction of target horizon of the Member 4 of Xujiahe Formation in the Fenggu Structure.