摘要
可控源电磁(CSEM)数据常常受到强人文噪声的污染,极大地影响了可控源电磁勘探的分辨率.为提高CSEM数据质量,本文提出一种基于支持向量机(SVM)的CSEM数据智能识别方法(CEEMD SVM方法),以代替传统的基于人工设定阈值的数据挑选方法.首先,通过互补集合经验模态分解(CEEMD)算法去除基线漂移噪声;然后,利用SVM对去除基线漂移后的数据进行智能识别,挑选出高质量信号.为验证该方法的有效性,首先进行了合成数据分析,然后将所提方法应用于广域电磁实测数据的处理.结果表明:SVM的平均识别准确率在92.00%以上;经过CEEMD SVM方法处理后,视电阻率由处理前的跳变形态变为连续光滑状态.
Controlled-source electromagnetic method(CSEM)signals are often contaminated by strong human noise,and the resolution of CSEM exploration is greatly affected.In order to improve the quality of CSEM data,in this paper,a new intelligent data processing method is proposed based on support vector machine(SVM)algorithm.Firstly,the baseline-drift noise is removed by complementary ensemble empirical mode decomposition(CEEMD)algorithm,and then the data is classified by SVM to select high-quality signals.In order to validate the proposed method,a targeted experiment was conducted using simulated noise and measured high-quality data,and then the method was applied to the measured data by the wide-field electromagnetic method(WFEM).The results show that the recognition accuracy of SVM is greater than 92.00%.After the treatment of the proposed method,the apparent resistivity changes from jumping shape to continuous and smooth.
作者
李广
丁迪
石福升
邓居智
肖晓
陈辉
何柱石
桂团福
Li Guang;Ding Di;Shi Fusheng;Deng Juzhi;Xiao Xiao;Chen Hui;He Zhushi;Gui Tuanfu(Engineering Research Center for Seismic Disaster Prevention and Engineering Geological Disaster Detection of Jiangxi Province(East China University of Technology),Nanchang 330013,China;Key Laboratory of Metallogenic Prediction of Nonferrous Metals and Geological Environment Monitoring(Central South University),Ministry of Education,Changsha410083,China)
出处
《吉林大学学报（地球科学版）》
CAS
CSCD
北大核心
2022年第3期725-736,共12页
Journal of Jilin University：Earth Science Edition
基金
国家自然科学基金项目(41904076,41830107,42074087,42130811)
中国博士后科学基金项目(2021M692987)
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(SDGD202008)
有色金属成矿预测与地质环境监测教育部重点实验室开放基金(2021YSJS02)。
关键词
广域电磁法
互补集合经验模态分解
信噪识别
支持向量机
可控源电磁法
wide-field electromagnetic method(WFEM)
complementary ensemble empirical mode decomposition(CEEMD)
signal-noise identification
support vector machine(SVM)
controlled-source electromagnetic method(CSEM)
