基于自组织神经网络在油气分层中的研究

测井曲线的分层研究是勘探和开发油气资源的重要手段,也是认识油气层的地质面貌,以及剖析油气藏量内在规律的一种有力武器。该文介绍了一种基于自组织神经网络对测井曲线进行聚类自动分层的识别方法,它是一种通过网络自身的调节,从而对输入数据进行聚类的方法。该文采用某地区的油气层数据来建立网络模型,首先采用了插值的方法,消除该测井数据随机干扰带来的噪声,同时保留了数据的完整性和代表性;然后通过利用自组织神经网络算法,对该数据自动进行四层的识别分类;最后结合人工分层的结果进行验证,以保证利用自组织神经网络识别的结果更加客观和可靠。该方法的可操作性强,原理简单易于实现,说明该算法对研究测井曲线具备一定的有效性和可行性。

Research on Oil and Gas Stratification Based on Self-organizing Neural Network

The logging curve stratification is an important means of oil and gas exploration and development,and also a powerful weapon to understand the geological features of oil and gas and to dissect the internal laws of oil and gas reserves.We introduce a recognition method based on self-organizing neural network for clustering and automatic stratification of logging curves,which is a method of clustering input data by adjusting the network itself.The data of oil and gas reservoir in a certain area is used to establish the network model.Firstly,the interpolation method is used to eliminate the noise caused by the random interference of the logging data,while preserving the integrity and representativeness of the data.Secondly,the self-organizing neural network algorithm is used to automatically identify and classify the data in four layers.Finally,the results of artificial stratification are verified to ensure the recognition results by self-organizing neural network are more objective and reliable.The method has strong operability,simple principle and easy implementation,which shows that it is effective and feasible for the study of logging curves.