Logging identification of low resistance oil reservoir and water-flooded layer based on grey relational analysis-extreme learning machine:Taking a case study of Guantao Formation in P block of Bohai Sea
After 20 years of development,P block of Bohai Sea has entered a high water cut stage,and a large number of low resistivity oil layers and water-flooded layers developed in Guantao Formation have no obvious difference in logging curve shape.In order to accurately identify water-flooded layer and classify water-flooded layer,machine learning algorithm is adopted in this paper.First,the sensitive parameter curve of the identification of water-flooded layer of the low resistivity oil layer is screened by using grey relational analysis;Then,the identification model of water-flooded layer of the extreme learning machine is constructed,and the model is trained to obtain the optimal parameters.It is applied to the actual data processing,the results show that the logging identification method of low resistance oil reservoir and water-flooded layer based on grey relational analysis and extreme learning machine has high prediction accuracy,and the coincidence rate is 89.3％,which is far better than the prediction results without grey relational analysis,and has practical application value.
万方数据
