Research On Intelligent Algorithm Of The Reservoir Sand Body Tracking And The Fault Identification Based On Well-seismic Information Fusion

Comprehensively taking advantage of characteristics of log data’s high longitudinalresolution and precise description of reservoir quality, and seismic data’s three-dimensionalreflection of reservoir’s structural characteristics and good plane connection, targeted attypical problems of detail correlation of reservoir bed, sand body trace and fault recognitionof reservoir geological study, based on multi-disciplinary data and materials of well logging,seismology and geology, this paper has studied intelligent analysis model and solutionalgorithm of logging and seismic information fusion, and also conducted design developmentand practical application of software prototype system.This paper has analyzed the cutting-edge theory and technology of detailed reservoirdescription based on logging and seismic information fusion, and also put forward researchmethods and technical frameworks of related contents to the subject. In the studies ofreservoir multi-disciplinary information fusion model and data management model, accordingto subject target and data processing procedure, this paper has established informationconcurrent method of information and data including well logging, seismology and geology,as well as meta-mode and management model of logging-seismic information fusion reservoirdescription data. In the automatic detail correlation of reservoir bed, this paper has putforward one kind of generalized distance norm measuring the overall similarity of two wells’characteristics of logging curves, directly taken the original logging curves of adjacent twocontrastive wells as the characteristic variables of detail correlation of reservoir bed, so as toconstruct its fitness function with constraints and particle swarm optimization applied in theoptimization solution of this function, which has better adaptability for stratigraphiccorrelation problem in the aspect of mechanism. In the study of thin sand body tracealgorithm, taking section of seismic data volume of adjacent two wells as the imagebackground, taking seismic interpretation horizon data as the bounder framework, adoptingthe edge detection technology principle of image processing, this paper has created the sandbody automatic algorithm based on the combination of discrete process neural networks andChaos Genetic algorithm. Under the restraint of small layer formation’s comparative resultsand sand body connecting conditions, it has realized the automatic trace of thin-layer sandbody. In the study of fault recognition method, based on data of seismic variable density mapand logging information of subordinate wells, on the basis of image edge search, applying theAnt colony tracking algorithm based on gradient information, this paper has realized thefault’s automatic identification and morphological distribution tracking of seismic datavolume.Through the process modeling of each research content and integration of informationintegrity, based on multidisciplinary data of logging and seismology, under the guidance ofreservoir evaluation theory and methods, this paper has set up intelligent algorithm and application technology applied in detail correlation of reservoir bed, sand body trace and faultrecognition. Based on CIFLog software platform of Chinese oil well logging, it hasdeveloped system integration and implementation of algorithm software. And also it hasacquired better results in the processing and application of practical data in work area of oilfield.