利用神经网络预测油气层损害

APPLICATION OF NEURAL NETWORKS IN THE PREDICTION OF FORMATION DAMAGE

利用人工神经网络预测油气层敏感性参数，根据大港油田的油气层岩石物性参数、岩石学分析参数以及油气层岩石敏感性参数等历史资料，训练神经网络，以此神经网络预测某些油气层的敏感性参数，该方法能迅速、准确地预测未来指定油气层的敏感性参数，为保护油气层提供依据。

Formation damage can decrease the output of oil, and leads to the huge economic loss.Sensitivity of formation is an important factor affectlng formation damage. Neural networks was applied to the prediction of formation sensitivity. The neural networks was trained with the information of rock physical properties petrology and rock sensitivity from Dagang oilfield. Formation sensitivity can be predicted by the trained neural networks. It is an effective method to predict the formation sensitivity by means of neural networks in Dagang oil field.