Logging curve reconstruction method based on genetic neural network algorithm

Abstract

The invention relates to the field of geophysical data processing, in particular to a logging curve reconstruction method based on a genetic neural network algorithm. The method mainly comprises the following steps of: 1, standardizing a logging curve; standardizing a conventional logging curve to be unified to the same dimension level; 2, establishing a neural network structure and training a network; determining neural network input and output, establishing a neural network structure according to the number of network layers, and training the network; 3, performing genetic manipulation; calculating a training error and a fitness function, and optimizing a network structure and a weight threshold by utilizing three genetic operators of selection, crossover and variation; and 4, performingwell logging curve reconstruction by utilizing the optimized network structure until the precision requirement is met, and outputting a result. Compared with a traditional well logging curve reconstruction method, the method has higher precision, the production cost can be reduced, the operation efficiency is improved, and the well logging curve reconstruction effect is improved.

Description

technical field [0001] The invention belongs to the technical field of geophysical data processing, and in particular relates to a logging curve reconstruction method based on a genetic neural network algorithm. Background technique [0002] In the field of mineral exploration, it is necessary to measure multiple different types of well logging curves to reduce the ambiguity of geological interpretation. [0003] However, in practice, it is often encountered that the logging curves are distorted or missing, for example, the diameter of the borehole is enlarged or the instrument is stuck, which makes some logging curves distorted or missing. [0004] Due to incomplete logging methods in early drilling, some important logging curves, such as sonic logging, may also be missing, which will bring difficulties to geological research. [0005] Therefore, how to accurately reconstruct these missing or distorted logging data is a problem worth exploring. For the reconstruction of w...

AI Technical Summary

Problems solved by technology

[0006]However, the traditional BP neural network has its own shortcomings, such as optimization is easy to fall into local minimum, etc., resulting in poor logging curve reconstruction accuracy (Zheng Qingsheng and Han Dakuang, 2007)

Images