特种油气藏2013，Vol.20Issue(2)：43-47,5.DOI:10.3969/j.issn.1006-6535.2013.02.010
利用概率神经网络预测成岩相——以鄂尔多斯盆地合水地区延长组长8段储层为例
Prediction of Diagenetic Facies with Probabilistic Neural Network —Taking Member Chang 8 of Heshui Area in Ordos Basin as an Example
摘要
Abstract
In view of the limited application of conventional diagenetic facies methods in Member Chang 8 of Heshui area of Ordos Basin due to its strong reservoir heterogeneity,a new method, namely, using probabilistic neural network to predict diagenetic facies, is proposed. Firstly, input parameters are studied by picking out sedimentary micro - facies and the data in logging curves - GR, SP, CAL, AC, CNL and DEN as input item parameters. Secondly, the probabilistic neural network is trained and tested. Finally, the probabilistic neural network that has been established is used to predict diagenetic facies in the study area. As a result, the accuracy of the method is over 90%. This method is applicable for the diagenetic facies study on the area where core is not sampled.关键词
概率神经网络/成岩相/长8储层/合水地区/鄂尔多斯盆地Key words
probabilistic neural network/ diagenetic facies/ reservoir Chang 8/ Heshui area/ Erdos Basin分类
能源科技引用本文复制引用
庞国印,唐俊,王琪,马晓峰,廖朋..利用概率神经网络预测成岩相——以鄂尔多斯盆地合水地区延长组长8段储层为例[J].特种油气藏,2013,20(2):43-47,5.基金项目
中国科学院"西部之光"联合学者项目"鄂尔多斯盆地延长组长8储层特征及其控制因素研究"(Y133WQ1-WQ) （Y133WQ1-WQ）
