| In this paper, following the basic principle of seismic attributes, we make a systematic exposition of the development process, current situation and development trend of the seismic attributes, and make a concrete analysis and study for the classification of seismic attributes, extraction methods, optimization attributes and making specific relationship building of reservoir and attribute, etc. At the same time, based on the seismic facies analysis, the preliminary studies are made.On the basis of theory, in BeiErXi District of Daqing Placanticline, for example, seismic attribute analysis techniques will be applied to the reservoir prediction. Through the fine interpretation of S21, S29 and S31 layer, we could determine the purpose layers and get the structural characteristics. The purpose layers are classified with the interval of S21-S29 and S29-S31, the intervals are extracted from more than 20 species of seismic attributes including root mean square amplitude, azimuth, frequency, etc and carrying on filter and poor normalizing by using Landmark and Stratimagic software. Attribute optimization programs are written to optimize five properties in each interval by using cluster analysis and correlation analysis, and to predict sand thickness and porosity in well points by using neural network, multiple regression and polynomial regression. Through statistical analysis of forecasting errors, we select the polynomial regression to predict the whole region laterally, then get the sand layer distribution and pore situation of the purpose layers. By using principal component analysis (PCA) and neural network, seismic waves of S21-S29 layer is divided into 6 categories, S29-S31 layers divided into 7 categories to classify seismic facies. With structural features, sand distribution and porosity conditions, favorable reservoir is classified withⅠandⅡfavorable areas to provide technical support for oil exploration and development. |
PDF Full Text Request