Research And Application Of Seismic Attribute Analysis For Key Parameters In Reservoir Development

With the deep development of oil and gas exploration,it is of far-reaching significance to estimate the spatial distribution of reservoirs and their parameters accurately.Since seismic attributes contain effective stratigraphic information,a multi-attribute comprehensive analysis method with seismic attribute optimization and key reservoir parameters estimation as the core is widely used.There are many types of seismic attributes,and a reasonable combination of seismic attributes can effectively reflect reservoir information and avoid problems such as computational difficulties caused by redundancy.In this paper,the “stepwise optimal”method is used to optimize seismic attributes.The minimum verification error is used as the criterion to avoid overfitting,and select the most suitable and sensitive seismic attributes subset at the same time.In the part of estimation of key reservoir parameters,this paper uses the seismic attribute analysis method based on the committee model.Support Vector Regression,Adaptive Network-based Fuzzy Inference and Wavelet Neural Network are used to make estimation respectively,and then sum the estimation results of the three algorithms by weight to build the committee model.In the above process,Particle Swarm Optimization algorithm is introduced and improved,which can optimize network configuration and improve the speed and stability.This paper conducts experiments about rock physics models and actual logging data to verify the “stepwise optimal” method and seismic attributes analysis methods based on the committee model.The results show that the two methods can indeed improve the applicability of algorithms and the accuracy of estimation results,while the positive effect of the improved Particle Swarm Optimization algorithm on the convergence speed and accuracy of estimation results has also been verified.Finally,the“stepwise optimal” method and seismic attributes analysis methods based on the committee model are applied to the second area of Gangdong in Dagang oilfield to obtain the inversion data volume of key reservoir parameters,and the relevant reservoir analysis is conducted.