摘要
切除废道(即道编辑)的工作是地震信号处理中重要而又工作量巨大的一步.本文提出了一种用人工神经网络实现的自动切除方法,所用的神经网络模型为新奇滤波器模型,方法简单实用,切除效果令人满意.文中介绍了方法的原理并给出了几个实验结果.
Noise is a difficult problem in seismic data Those traces with dominating noise should be removed before processing. This editing work is labour consuming if done manually and is difficult for automation because the noises are of many different kinds. This paper provides an artificial neural network approach doing the automatic editing task. The neural model used is the Novelty Filter model. Using this approach, noisy trtces can be easily distincted from good ones disregarding their different kinds. Experments show that this approach is simple and efficient and the results are acceptable.
出处
《地球物理学报》
SCIE
EI
CAS
CSCD
北大核心
1992年第5期637-643,共7页
Chinese Journal of Geophysics
基金
国家自然科学基金
关键词
地震
信号处理
滤波器
人工智能
Seismic signal processing, Automatic editing, Artificial neural networks,Novelty filter
