摘要
自然电位信号来源复杂,实测数据由多种自然电流源叠加贡献,直接反演可能获得假异常.本文利用基于最小相关约束的非负矩阵分解算法开展混合自然电位信号盲源分离预处理以提取目标自然电位信号,然后采用基于电阻率约束的阻尼粒子群优化算法实现自然电流源三维反演成像,并通过聚焦于金属矿产勘探领域的数值算例、物理模拟实验以及现场案例测试了算法的有效性.研究表明,非负矩阵分解预处理能大幅提高目标自然电位信号的信噪比,为反演提供优质的数据基础,提升数据解释的准确性;基于电阻率约束的粒子群优化算法汇集了梯度反演及全局优化反演的优点:不依赖于初始模型、具有全局搜索能力、场源分布符合实际物理意义.本文研究对金属矿产勘探领域实测自然电位数据预处理及反演解释具有参考价值.
The origins of self-potential(SP)signals are complex,as the measured data result from multiple SP current sources.This complexity can lead to false anomalies when directly conducting data inversion.In this paper,we employed a non-negative matrix decomposition algorithm constrained by the least correlated component to perform blind source separation,effectively extracting target SP signals for mixed SP data.Subsequently,we successfully achieved 3D inversion imaging of SP current sources using a damped particle swarm optimization algorithm incorporating resistivity constraints.We evaluated the effectiveness of our approach through numerical examples,physical simulation experiments,and a field case study focused on metal mineral exploration.The results demonstrate that non-negative matrix decomposition preprocessing significantly enhances the signal-to-noise ratio of the target SP signals,thereby providing a high-quality data foundation for inversion and improving data interpretation accuracy.The resistivity constraint-based damped particle swarm optimization algorithm leverages the strengths of both gradient inversion and global optimization,eliminating dependence on the initial model,while facilitating global search capability and ensuring that the SP source distribution has meaningful physical interpretations.This research offers valuable insights for the preprocessing and inversion of real measured SP data in the field of metal mineral exploration.
作者
崔益安
李浩
王璞
谢静
张丽娟
罗议建
CUI YiAn;LI Hao;WANG Pu;XIE Jing;ZHANG LiJuan;LUO YiJian(School of Geosciences and Info Physics,Central South University,Changsha 410083,China;Department of Geophysics,School of Earth and Space Sciences,Peking University,Beijing 100871,China;School of Architecture and Electrical Engineering,Hezhou University,Hezhou Guangxi 542899,China;China Railway Design Corporation,Tianjin 300308,China)
出处
《地球物理学报》
北大核心
2025年第9期3600-3615,共16页
Chinese Journal of Geophysics
基金
国家自然科学基金(42174170,42204135,42474173)
中国博士后科学基金(2024M750105)资助.
关键词
自然电位
非负矩阵分解
电阻率约束
粒子群优化反演
金属矿产勘探
Self-potential
Non-negative matrix decomposition
Resistivity constraint
Particle swarm optimization inversion
Metal mineral exploration
