节点文献
用人工神经网络法预测气-水边界
PREDICTION OF THE GAS-WATER SURFACE BY AN ARTIFICIAL NEURAL NETWORK METHOD
【摘要】 文章研究了在构造复杂地区(如高陡构造带、断层发育等),地震记录上储层的信噪比低的情况下,用人工神经网络方法预测气水边界。对影响预测效果的某些因素作了试验,其中包括地震特征参数的选择、计算时窗大小、数据圆滑方式、学习样本与预测区域的选择等。文中还给出了已知剖面的预测结果,并被钻井证实是成功的。
【Abstract】 An artificial neural network method used to predict the natural gas-water surrface is discussed in this paper. In the study area,the geologic structure is complicated with a steep anticline andthe signal-noise ratio of the seismic reflection data on the reservoir beds is rather low. Experimentshave been made with some factors influencing the predicting effects which include the choice of characteristic parameters,the time window of the computation,the smooth method of the seismic data aswell as the choice of the samples for training network and the region predicted. The predicted results oftwo profiles in the study area are given and proved to be correct by well records.
- 【文献出处】 成都理工学院学报 ,JOURNAL OF CHENGDU UNIVERSITY OF TECHNOLOGY , 编辑部邮箱 ,1995年03期
- 【分类号】P618.130.8
- 【下载频次】42