摘要
提出了一种改进的阿基米德优化算法(Improved Archimedes Optimization Algorithm,IAOA),旨在提高地震波阻抗反演的精度和稳定性,以解决地下介质中岩性分析和存储计算等问题。IAOA在AOA的基础上优化了种群初始化,在迭代过程中加入了能够自动平衡全局搜索和局部开发性能的能力,丰富了AOA迭代后期的种群信息。为了测试IAOA用于地震波阻抗反演的性能,设计了3个理论地质模型用于算法的波阻抗反演,并且将IAOA、AOA和粒子群优化算法(Particle Swarm Optimization,PSO)分别反演同一个模型以对比3个算法的性能,选择PSO作为对比算法是因为其在地震波阻抗反演中广泛应用且性能稳定。模型测试结果表明,IAOA的反演结果精度高,具有较稳的收敛过程,可有效地用于波阻抗的反演。最后,使用了一个来自中国北部松辽盆地某研究区内的实测数据来测试IAOA反演实际数据的能力。反演结果表明,IAOA的实用性强,能获取有效的地下地层信息,有利于后续储层计算等研究的开展。
The article proposes an Improved Archimedes Optimization Algorithm(IAOA)aimed at enhancing the accuracy and stability of seismic impedance inversion to address issues such as lithological analysis and storage calculations in subsurface media.IAOA improves population initialization based on AOA,incorporates automatic balancing of local and global development during iteration,and enhances population information in the later stages of AOA iteration.To evaluate the performance of IAOA for seismic wave impedance inversion,three theoretical geological models were designed for algorithm testing.The results were compared among IAOA,AOA,and Particle Swarm Optimization algorithm(PSO)when applied to invert the same model.The PSO algorithm was selected as a comparison because of its wide application and stable performance in seismic wave impedance inversion.The model test results demonstrate that IAOA achieves high precision in inversion results,maintains a stable convergence process,and can effectively contribute to wave impedance inversion tasks.Finally,this paper employs measured a research area in northern China to assess the ability of IAOA in inverting actual data.The obtained results indicate that IAOA possesses strong practicality by providing effective subsurface stratigraphic information beneficial for subsequent research such as reservoir calculation.
作者
姜茂燃
姚振岸
毕升博
丁文蔷
JIANG Maoran;YAO Zhenan;BI Shengbo;DING Wenqiang(Jiangxi Provincial Earthquake Disaster Mitigation and Engineering Geological Hazard Detection Engineering Research Center(East China University of Technology),330013,Nanchang,PRC)
出处
《江西科学》
2025年第4期660-667,共8页
Jiangxi Science
基金
国家自然科学基金青年项目(42004113)
江西省自然科学基金项目(20212BAB211003)
江西省教育厅科学技术项目(GJJ200728)
江西省防震减灾与工程地质灾害探测工程研究中心开放基金项目(SDGD202006)
东华理工大学研究生创新专项资金项目(DHYC-202417)。
关键词
波阻抗
地震数据
全局非线性优化
阿基米德优化算法
wave impedance
seismic records
global nonlinear optimization
Archimedes Optimization Algorithm
作者简介
姜茂燃(1998-),男,硕士研究生,主要研究方向为基于深度学习的地震数据重建方法和波阻抗反演技术;通信作者:姚振岸(1991-),男,博士,副教授,硕士生导师,主要从事地震勘探科研与教学工作。E-mail:an6428060@163.com。
