赵平起^1,,
何书梅¹,
倪天禄¹,
赵明¹,
张家良¹,
吴吉忠^1,2,
魏朋朋¹,
李闻达^3,4,5,6,
白文磊^3,4,5,6,,
1. 中国石油大港油田分公司, 天津 300280
2. 东北石油大学, 大庆 163318
3. 中国科学院地质与地球物理研究所, 北京 100029
4. 中国科学院地球科学研究院, 北京 100029
5. 中国科学院油气资源研究院重点实验室, 北京 100029
6. 中国科学院大学, 北京 100049

基金项目: 港东二区油藏渗流地球物理技术提高采收率研究"标段1:地震波场与油藏流场耦合关系研究"(FRSG), 国家自然科学基金重点项目(41630319), 国家科技部重大科学仪器设备开发专项重点研发项目(2018YFF01013500-03)和国家科技重大专项课题(2016ZX05024-001-001)联合资助


详细信息

作者简介: 赵平起, 1964年生, 教授级高级工程师, 1985年获华东石油学院采油工程专业学士学位, 2001年获成都理工大学油气田开发工程硕士学位, 2004年获中国科学院地质学博士学位, 现任中国石油大港油田公司副总经理, 主要从事油气田开发管理及研究.E-mail: zhaopq1203@163.com

通讯作者: 白文磊, 1995年生, 硕士, 主要从事高铁地震学、复杂介质波场数值模拟与成像、智能算法及应用研究.E-mail: wl_bai@163.com

中图分类号: P631

收稿日期:2020-12-03
修回日期:2021-06-08
上线日期:2021-07-10

Numerical modelling of elastic waves based on the BFO-PSO algorithm

ZHAO PingQi^1,,
HE ShuMei¹,
NI TianLu¹,
ZHAO Ming¹,
ZHANG JiaLiang¹,
WU JiZhong^1,2,
WEI PengPeng¹,
LI WenDa^3,4,5,6,
BAI WenLei^3,4,5,6,,
1. PetroChina Dagang Oilfield Company, Tianjin 300280, China
2. Northeast Petroleum University, Daqing 163318, China
3. Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100029, China
4. Innovation Academy for Earth Sciences, Chinese Academy of Sciences, Beijing 100029, China
5. Key Laboratory of Petroleum Resource Research, Chinese Academy of Sciences, Beijing 100029, China
6. Chinese Academy of Sciences, Beijing 100049, China


More Information

Corresponding author: BAI WenLei,E-mail:wl_bai@163.com

MSC: P631

--> Received Date: 03 December 2020
Revised Date: 08 June 2021
Available Online: 10 July 2021


摘要
摘要:地震波正演模拟是地震反演与成像的基础和关键，有限差分算法广泛应用于地震波数值模拟，差分算子的精度直接影响数值模拟的质量和效率.本文提出一种BFO-PSO算法下的有限差分算子优化方法，并应用其进行弹性波数值模拟.首先，将BFO算法中的趋化、复制、驱散三个步骤引入PSO算法，形成具有更好全局搜索能力和更快收敛速度的BFO-PSO混合优化算法；之后构造包含有限差分系数的目标函数，并应用BFO-PSO混合优化算法求取最优解，获得优化的有限差分算子；最后应用此优化的有限差分算子在不同模型上进行弹性波数值模拟.根据频散曲线及数值模拟结果，可以分析得出，BFO-PSO算法优化后的有限差分算子在保证计算效率的同时，具有更高的精度，可以有效压制数值频散，提高数值模拟的精度和效率.
关键词: BFO-PSO算法/
有限差分/
数值模拟/
弹性波

Abstract:Forward modelling of elastic waves is the foundation and key of seismic inversion and imaging. The Finite-Difference (FD) method is one of the most popular means used in such numerical modelling, in which the accuracy of the difference operator affects the quality and efficiency of the modelling. In this paper, we propose an optimal FD method based on the BFO-PSO algorithm, and apply it to numerical modelling of elastic waves. First, the three steps of chemotaxis, replication and dispersal in the BFO algorithm are introduced into the PSO algorithm, and a new BFO-PSO hybrid optimization algorithm with better global searching ability and faster convergence speed is generated. Then, we construct the objective function containing the FD coefficients. The BFO-PSO algorithm is used to solve the object function and obtain the optimal FD method. Finally, numerical modelling of elastic waves is performed with this optimal FD method based on the BFO-PSO algorithm on some models. Numerical dispersion analysis and modelling results indicate that the optimal FD method based on the BFO-PSO algorithm has a high accuracy while ensuring calculation efficiency, and can efficiently suppress the numerical dispersion and improve the accuracy and efficiency of numerical modelling.
Key words:BFO-PSO algorithm/
Finite-difference/
Numerical modelling/
Elastic waves



PDF全文下载地址:

http://www.geophy.cn/data/article/export-pdf?id=b22a183d-9158-4485-807b-2091c26dea63