The Application Of Log Parameters Inversion Under The Seismic Constrains

As oil and gas exploration are more difficult and complex than that before. Usingseismic inversion to extract the hidden information from seismic data can improve theaccuracy of reservoir prediction, and has been become one of the effective means ofreservoir prediction. With decades of development, the application of seismic inversiontechnology has formed certain process and approaches. Because of the complicatedrelationship between seismic attributes and reservoir parameters, it is difficult to com-plete reservoir prediction and interpretation task by using the seismic attributes alone.It is necessary to comprehensive interpretation with seismic inversion analysis, loginformation and geological information. Therefore, seismic attributes analysis tech-niques which are used to predict reservoir lithology and physical properties are stillneeded to continue developing and researching.Traditional seismic inversion is mainly based on the model. As input of inversion,the hypothetical model is built by man-made. The model needs to be close to the realgeological structure of subsurface. It is very difficult to using insert and extrapolationmethods and multiple iterative inversion techniques to get close results to the truesolution. This paper tries to choose the sensitive logging curves to the reservoir li-thology, physical properties, oiliness and to get3D chosen logging data volume byartificial intelligence neural network technology under the seismic attributes constrains.The results are the characterization of lithology and physical properties of3D datavolume, and are important to predict reservoir lithology and physical property, etc.This paper studies the principles of log parameters inversion under the seismicconstrains and its application in reservoir prediction. Through testing different sensitiveparameters for prediction of the reservoir, the author summarizes the workflow ofseismic logging parameters inversion under seismic constrains. This method is appliedto the reservoir prediction in XinJiang. The application of this method gets much betterresults than other normal way, and provides the support for reservoir prediction in thestudy area.