Study On Artificial Neural Networks' Application In Thin And Poor Water-Flooded Layer Interpretation

The brunt blocks of Daqing Placanticline have been in High Water-Cut Period. All thick oil zones,which are oil zones of high porosity and saturation are water-breakthrough. Oil displacement efficiencies are reducing obviously. Thin and poor water-flooded layers have become main goal of increasing reserves and raising production.Now, Thin and poor water-flooded layers interpretation is focus and difficuly of oil-field exploration. Many problems have to be solved. In this paper the author has developed some researching work against disvantages of common interpretation. The paper includes three aspects. First,well-logging data from Saertu,Putaohua and Gaotaizi oilfield has been preprocessed and standardize. Second,Qualitative model of forecasting water-flooded layers interpretation and quantitative model of forecasting the parameters such as porosity and saturation have been established using artificial neural network. Last, the initial interconnecting weights and thresholds of BP ANN have been optimized using genetic algorithms based on real number coding and optimized forecasting models have been established respectively. So using the model and well-logging data from Saertu,Putaohua and Gaotaizi oilfield water-flooded layers have been identitied qualitatively, and the parameters such as porosity and saturation have been forecasted quantitatively. Then the results have been analyzed and evaluated.Studying relative theories and analyzing old interpretation methods for a long time the author exploited a mathematical model identifying water-flooded layers using artificial neural network. This model can identify thin and poor water-flooded layers qualitatively and can forecast the parameters such as porosity and saturation quantitatively. By optimizing the initial interconnecting weights and thresholds of BP ANN using genetic algorithms the model has been impoved. It increases qualification rate and make a contribution to raising profit and techlonogy progress of of our company.