Research On Intelligent Evaluation Method Of Water Flooded Layer Based On Comprehensive Logging Data

With the development of the oil and gas exploration,due to the use of the flood,most of the fields have been in the high water level,and they have made a higher and higher demand for watered-out evaluation.The main task of logging evaluation is to obtain information that can reflect underground geological conditions and to judge the oil and gas situation in the underground.Nowadays,logging evaluation has been developed in the direction of quantitative evaluation with the help of computer technology.The evaluation method of logging water flooded layer has been turned from single parameter to multi parameter,which has achieved good results.The logging technology has the condition of multi-parameter evaluation,and the logging technique is based on the analysis of stratigraphic rock samples,which is not affected by the heterogeneity of strata,and has obvious advantages in the evaluation of watered-out layer,and has achieved certain effect in practical application.For a long time,the professional personnel in logging field have accumulated the evaluation theory and analysis experience.Logging data and computer technology combine to carry out the evaluation of water flooded layer has important research significance,and it has a positive effect in the development of logging field and oil and gas exploration.First,obtains from the logging technology,and expounds the analysis principle and parameters of logging technology,analyzes the logging technology of water flooded layer evaluation method.Through the study of water flooding,the characteristics of the saturated hydrocarbon gas chromatography are analyzed,and it analyzes the change of water rate in saturated hydrocarbon gas chromatography.Using the characteristics of wavelet packet multi-level decomposition,the characteristics of multi-level decomposition of wavelet packets are presented,and a method for the extraction of saturated hydrocarbon gas chromatographic features by wavelet packet analysis is presented.Secondly,the basic structure and training algorithm of wavelet neural network are studied.Because of the BP training algorithm,the wavelet neural network is deficient in the convergence ability.Therefore,the particle swarm optimization algorithm is introduced,and the basic principle and algorithm flow of particle swarm optimization algorithm is analyzed,and the evaluation model of wavelet neural network water flooding layer is proposed.The wavelet neural network water flooded layer evaluation model is proposed,and the particle swarm algorithm is used as the training algorithm of wavelet neural network,which can optimize the convergence speed and searching ability of wavelet neural network,and construct a sample set for training of network model.Finally,based on the above theory and technology research,design and implement a comprehensive logging water flooded layer intelligent evaluation system.It has realized the function of chromatographic modeling,comprehensive schematic modeling,comprehensive chart interpretation data processing and comprehensive chart interpretation.The system can deal with that characteristic of the chromatographic features,establish the water flooded evaluation standard according to the comprehensive logging parameter,and can perform the water flooding layer evaluation on the unknown reservoir.