SVM Vp/Vs prediction method based on small sample machine learning

Abstract

The invention discloses an SVM Vp / Vs prediction method based on small sample machine learning in the crossing field of machine learning and geophysical prediction. According to the method, firstly, wave impedance Ip and Vp curve data of at least one well and label data Vp / Vs (dipole acoustic wave data is needed) are utilized to iteratively train a small sample machine learning SVM model, and the required precision is achieved; and then the trained SVM Vp / Vs prediction model is applied to other dipole-free acoustic well Vp / Vs prediction. Vp / Vs curve prediction can be carried out on the dipole-free sound wave data well by using the method, the method is rapid and simple, and the working efficiency is improved by more than 10 times compared with that of a traditional method. The method is particularly suitable for areas with less drilling wells.

Description

technical field [0001] The invention belongs to the cross field of machine learning and geophysical prediction. Background technique [0002] The velocity ratio Vp / Vs of compressional and shear waves is an important petrophysical parameter and an important parameter for calculating Poisson’s ratio (Poisson’s ratio is a nonlinear function of Vp / Vs). The larger the Vp / Vs, the larger the Poisson’s ratio, so Vp / Vs is called pseudo-Poisson's ratio. Poisson's ratio or Vp / Vs plays an important role in reservoir prediction and hydrocarbon detection in oil and gas exploration and development. Usually in clastic rock formations, sandstones exhibit low Poisson’s ratio or low Vp / Vs characteristics, while mudstones, including coal seams, exhibit high Poisson’s ratios or high Vp / Vs characteristics, so Poisson’s ratio can be used to distinguish sandstone and mudstone . If oil and gas are stored in sandstone, especially gas reservoirs, Poisson's ratio is lower, so low Poisson's ratio is ...

AI Technical Summary

Problems solved by technology

[0004] (1) The traditional petrophysical modeling method is very cumbersome and requires professional geophysical logging engineers to master this method;

[0005] (2) Traditional petrophysical modeling methods require a large number of parameters, including rock elastic modulus, reservoir parameters, and reservoir parameters, etc., but these parameters are usually difficult to obtain;

[0006] (3) The traditional petrophysical modeling method has a large workload and low efficiency

Images