西南石油大学学报(自然科学版) ›› 2008, Vol. 30 ›› Issue (1): 63-66.DOI: 10.3863/j.issn.1000-2634.2008.01.017

• 地质勘探 • Previous Articles     Next Articles

Application of BP Neural Network Technology in Sonic Log Data Rebuilding

YANG Zhi-li 1,2,ZHOU Lu 2,PENG Wen-li 3,ZHENG Jin-yun 2   

  1. 1.Hangzhou Research Institute of Petroleum Geology,Hangzhou Zhejiang 310023,China;2.Resources and Environment institute,Southwest petroleum University,Chengdu Sichuan 610500,China;3.Research Institute of Exploration and Development,Xinjiang Oil field Company,Karamay Xinjiang 834000,China
  • Received:2006-01-06 Revised:1900-01-01 Online:2008-02-20 Published:2008-02-20

PDF (PC)

737

Abstract: The log reconstruction method based on neural network is a effective one to improve the quality of sonic log.The method,by means of non-linear algorithm of neural network technology,is used to set up a non-linear relation among sonic log curve,SP,Rt,Gr and curves.The sonic log curve is reconstructed by using BP neural network technology combining SP,Rt,Gr curves data as to enhance the information quality of curve reconstruction and ensure the accuracy of curve reconstruction.The practice in Shinan area,Junggar Basin proves the method can greatly improve the quality of sonic log curve where well wall is affected by collapse.

Key words: log reconstruction, neural network, sonioc log, Junggar Basin, Shinan area

CLC Number: