基于遗传神经网络算法的测井曲线重构技术  

Well logging curve reconstruction based on genetic neural network

作　　者：金永吉 张强[2] 王毛毛[2] JIN YongJi;ZHANG Qiang;WANG MaoMao(China National Uranium Co.,Ltd.,Beijing 100013,China;No.216 Geological Team,China National Nuclear Corporation,rümqi 830011,China)

机构地区：[1]中国铀业有限公司,北京100013 [2]核工业二一六大队,乌鲁木齐830011

出　　处：《地球物理学进展》2021年第3期1082-1087,共6页Progress in Geophysics

基　　金：新疆准噶尔盆地及周边地区铀矿资源调查评价与勘查(201906)资助。

摘　　要：本文使用遗传算法对传统神经网络的拓扑结构、权值和阈值进行优化,提出了基于遗传神经网络优化方法的测井曲线重构技术,可克服传统神经网络方法易陷入局部最小值的缺点.以声波测井曲线的重构为例,确定遗传神经网络有关参数并分析其重构效果.为了验证方法的有效性,分别对声波曲线、电阻率曲线和自然伽马曲线进行了重构,结果表明,遗传神经网络的曲线的重构结果要优于传统神经网络.因此,基于遗传神经网络的测井曲线重构具有较好的精度和实用性.A technique to reconstruct wireline logs based on the Genetic Neural Networks(GNN) optimization is presented in this paper. Using the genetic algorithm to optimize the traditional neural network’s topology structure, weight and threshold, it can effectively overcome traditional neural network’s shortcomings of redundancy structure, tendency to fall into local minimum, etc. Taking the reconstruction of acoustic logging curve as an example, the optimized structure and relevant parameters of GNN was determined, and the validity of the technique was demonstrated. Then, in order to verify its feasibility, more curves including acoustic, resistivity and Natural gamma logs were reconstructed. It shows that the GNN are superior to those by traditional neural network. Therefore, the logging curve reconstruction based on GNN has better accuracy and practicability.

关 键 词：测井曲线重构技术 神经网络 测井曲线 遗传算法 

分 类 号：P631[天文地球—地质矿产勘探]