摘要
层析成像的分辨率取决于对地震 (声波 )全波列记录信息提取的精度 ,本文阐述了从地震(声波 )全波列记录提取特征信息 ,建立三层神经元网络模型 ,利用神经元网络对地震 (声波 )全波列记录的信息提取 ,如初至波走时拾取。本文讨论了神经元网络模型的建立和神经元网络模型的学习。侧重讨论了初至波出现时的特征并对它进行了定量化描述 ,用它对地震 (声波 )全波列记录信息提取 。
The resolution radio of tomographic imaging depends on the acquisition precision of seismic (acoustic) full wave form information.The present paper deals with the collection of characteristic information from seismic (acoustic) full wave form records,the establishment of three wave form records,the establishment of three layer neuron network model,the application of neuron network to the information collection of seismic (acoustic) full wave form records,such as the traveltime acquisition of initial wave.The paper also discusses the establishment of neuron network model and the study of neuron network model,with an emphasis on the characteristics of the initial wave when it appears.A quantitative description is given,and its application to the information collection of seismic (acoustic) full wave form records has exhibited very high precision.
出处
《物探与化探》
CAS
CSCD
1998年第2期143-148,共6页
Geophysical and Geochemical Exploration
基金
地质行业科技发展基金资助
关键词
井中声波
神经网络
层析成像
全波列信息
borehole acoustic wave (seismic) perspective,neural network,tomographic imaging.
