Optimal retention strategy-based genetic algorithm wave impedance inversion method

Abstract

The invention provides an optimal retention strategy-based genetic algorithm wave impedance inversion method and belongs to the field of geophysical inversion, which particularly relates to a wave impedance inversion technology in the oil-gas geophysical exploration field. The invention provides an improved genetic algorithm-based wave impedance inversion method, wherein the early-maturing convergence problem of a standard genetic algorithm in wave impedance inversion is solved. As a result, an obtained inversion result is more reliable. The method mainly comprises the following steps of (1) constructing a target function of wave impedance inversion according to a convolution model; (2) estimating the seismic wavelet through the homomorphic theory; (3) coding the wave impedance in a binarycoding mode; (4) calculating the fitness value of each individual by using a target function, and carrying out quantitative evaluation on the individual; (5) generating a new-generation population according to the selection mode of the optimal reservation strategy; (6) according to designed crossover and mutation operators, carrying out the genetic operation; (7) converting individual genotypes into phenotypes according to corresponding decoding modes and realizing algorithm circulation; and (8) adopting a recursion method for calculating the wave impedance.

Description

technical field [0001] The invention provides a genetic algorithm wave impedance inversion method based on the optimal retention strategy, which belongs to the field of geophysical inversion, especially the wave impedance inversion technology in oil and gas geophysical exploration technology, and is used to solve the problem of the standard genetic algorithm in Premature convergence problem in wave impedance inversion. Background technique [0002] The standard genetic algorithm has the weakness of premature convergence and slow convergence speed in practical application. Most of the current improvement methods are aimed at the macroscopic operation of the population, genetic operation, genetic operator improvement and parallel operation. In terms of coding methods, dynamic coding methods, floating-point number coding methods, Gray coding methods, etc. have been developed. In terms of population macro operations, Yao Weifan (2015) proposed a dynamic niche co-evolution model....

AI Technical Summary

Problems solved by technology

In terms of coding methods, dynamic coding methods, floating-point number coding methods, Gray coding methods, etc. have been developed. In terms of population macro operations, Yao Weifan (2015) proposed a dynamic niche co-evolution model. The improvement of the selection operator On the one hand, Zhang Jing (2015) proposed an adaptive sorting selection method. In terms of crossover operator improvement, Davis (1991) proposed a serial number crossover operator and a uniform sorting crossover operator. In terms of mutation operator improvement, Liu Li (2015) proposed a gene adaptive mutation genetic operator. The above-mentioned improved algorithm improves the search performance of the genetic algorithm to varying degrees, but for the specific problems of wave impedance inversion, the above-mentioned improved measures The implementation process is cumbersome, the adaptability to different problems is poor, and it fails to show obvious advantages in terms of convergence

Images