Application Of Cellular Genetic - Neural Network Method For Oil And Gas Prediction

With much deeper understanding to geology and the application of new exploration technologies, It is very significance to discuss the problem that apply the method to predict gas reservoirs. It is a efficiency technology to predict the distribution of enrichment gas reservoirs by the nonlinear and routine method.In this paper, we extracts some features from seismic data using nonlinear method and linear method, and compressive these parameters, and we got less feature parameters. In this process, the nonlinear methods used to extracts feature parameters and design the machine classifier.The nonlinear feature parameters including :DIM,LYA,TU,PBX. And in this paper,CA,GA,ANN are used to design the machine classifier. And in this paper, linear method also used to extract feature parameters.In this paper, This method was used to part of the heavey detrital rock of the west of Sichuan data to predict the gas reservoirs, the results shows that this method has good effect.In this paper, we not only extract some linear feature parameters but also four nonlinear feature parameters, they are DIM,LYA,PBX,TU, It made the predict result more credibility. The concept of CA was also introduced when design the machine classifier, and at this process, we mesh the ANN ,GA and CA, introduce a mixed arithmetic, CAGA-ANN, which optimized the compute speed, and avoid the problem of low efficiency and local optimum of the BP arithmetic.