期刊文献+

基于支持向量机的岩石物理统计应用 被引量:1

Application of petrophysical statistics based on support vector machine
下载PDF
导出
摘要 针对岩石物理统计与波阻抗反演之间相对关系准确性、真实性不高等问题,采用支持向量机对岩石物理统计进行处理,得到真实和准确的数据。岩石物理统计是地质模型与地震响应的桥梁,其准确与否直接决定反演结果的精确度,对反演具有非常重要的意义。波阻抗是连接地球物理与地质的桥梁,在岩石物理统计中,波阻抗与岩石物性的相对关系就显得非常重要。通过支持向量机,对波阻抗与岩石物性的相对关系进行准确合理的分析,得到相对合理、真实的波阻抗和岩石物性相对关系,从而提高波阻抗反演的准确性和真实性。应用支持向量机对岩石物理统计数据进行处理,能够避免人为划分的误差,从而提高波阻抗反演结果的准确性,避免人为划分对结果的影响。 Aimed at the problems of relative accuracy and authenticity of rock physical statistics and wave impedance inversion, the sup-port vector machine was used to process the petrophysical statistics to obtain real and accurate data. The petrophysical statistics is a bridge between geological model and seismic response and the accuracy of petrophysical statistics directly determines the accuracy of in-version results, so the statistics of rock physics is very important for inversion. The wave impedance is a bridge between geophysics and geology,so the relative relationship between wave impedance and petrophysical property is very important in petrophysical statistics. The relative relationship between wave impedance and petrophysical properties is analyzed by support vector machine. The relative relation-ship between wave impedance and petrophysical property is obtained, which can provide accurate impedance and physical property for wave impedance inversion. So as to improve the accuracy and authenticity of wave impedance inversion. The application of SVM to the processing of petrophysical statistical data can avoid the error of artificial division, so as to improve the accuracy of wave impedance in-version and avoid the influence of artificial on the result.
作者 杨飞 邱鹏 刘金帅 Yang Fei Qiu Peng Liu Jinshuai(School of Geophysics & Oil Resources ,Yangtze University, Wuhan 430100, China)
出处 《能源与环保》 2017年第5期194-197,共4页 CHINA ENERGY AND ENVIRONMENTAL PROTECTION
关键词 支持向量机 岩石物理 波阻抗 地震波阻抗反演 分类间隔 support vector machine rock physics wave impedance seismic wave impedance inversion classification interval
  • 相关文献

参考文献6

二级参考文献26

共引文献156

同被引文献16

引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部