Research On Reservoir Image Fusion And Oil/gas Nonlinear Detection

On the basis of the theories and methods of the image fusion and nonlinearinversion algorithm, this paper developed new-style reservoir image fusion technique,seismic nonlinear oil/gas detection method and technique by using new methods,techniques and ideas.(1) Based on the theory and method of image fusion, this paper developedreservoir image fusion technique which is based on PCA transform and wavelettransform (including second era wavelet). The reservoir image fusion technique basedon PCA transform utilizes the dominant component focused the most energy toreconstruct the reservoir image; however, the reservoir image fusion technique based onwavelet transform utilizes the properties and details contained in each images toperform fusion in many decomposition layers and many frequency bands. So, thistechnique is a new oil/gas information integrative processing technique which canprovide more accurate, complete, and reliable image descriptions of the target zones inthe reservoirs; and also is a new technique which can truly describe the reservoir imageafter fusion, and reflect the same geologic factors in multiple seismic attributes orparameters, and directly transform the seismic attributes or parameters into thegeological information.(2) Based on the theory and method of the nonlinear algorithm, this paper tht firstdeveloped a new seismic nonlinear inversion method that integrates all advantages ofgenetic algorithm, chaos algorithm, ANFIS algorithm and tabu searching algorithm, thatis, a global optimization hybrid intelligent algorithm. This algorithm implementscomputation with probability searching mode, moreover, the probability changesself-adaptively to make the hybrid algorithm uniform. In this technique, the networkstructure with multi-inputs and multi-outputs are adopted, and under the constraint ofthe logging data, the self-adaptive weight function and comprehensive nonlinearmapping relationship are established and nonlinear mapping relationship areautomatically updated, which can constrain and control the inversion results, andimprove inversion precision and resolution greatly. Therefore, it is a seismic inversion method with high precision and high resolution, and can perform seismicmulti-attributes inversion.(3) Based on the elastic parameters is the optimal parameter for reservoir lithologyprediction and oil/gas detection, the paper developed elastic parameter inversiontechnique series, which consists of poisson'ratio inversion based on equation, elasticparameter conversion technique, improved weighting stack technique and P-wave andS-wave simultaneous elastic parameter inversion technique which based on hybridintelligent algorithm, and elastic parameter integrative interpretation method. Thistechnique series belongs to prestack inversion technique, and is an advanced inversiontechnique.(4) Reservoir density inversion method is based on hybrid intelligent seismicnonlinear inversion method, and applies seismic acoustic impedance data to performreservoir density inversion with the constraint of the logging density data, and gainsreservoir density section. Then, fluid density computation method is adopted to obtainthe fluid density and fluid density section, which are the optimal parameters for oil/gasdetection.(5) "Reservoir prediction and oil/gas detection software system" developed by theauthor has four functions, including reservoir image fusion function, velocity anddensity parameters etc. inversion function, various elastic parameters inversion functionand interpretation function. This software system has the graphical interfaces, isconvenient for operation, and has strong application values. This reservoir image fusionsoftware is the first one developed at home.(6) Seismic nonlinear inversion method was applied to obtain 3D velocity volumeand 3D density volume with high resolution in Fengle area of Daqing oil field, andperform the most favorable volcano rock body recognition and oil/gas prediction, andgood geologic results have been obtained. Therefore, this method provided new methodand technique for the recognition and prediction of the complex volcano rock bodies.Moreover, this technique was applied to successfully predict oil/gas accumulation zonesand their locations in the delineated reservoir target zones in Sichuan basin, andimprove the successful rate of the exploration and exploitation in this area greatly. Inconclusion, this technique and software system are suitable for various areas andreservoir types with different properties, and have high science value, application valueand popularizing application prospects.