基于神经网络的柳林地区煤储层渗透率预测方法

Permeability Prediction Method of Coal Reservoir in Liulin Area Based on Neural Network

分析了煤储层渗透率预测中存在的问题,提出基于测井信息的GA-BP神经网络预测煤储层渗透率方法,分析其机理及测井参数标准化处理方法。以柳林地区56口井的试井和测井资料为基础,利用灰色关联分析法优选6个测井参数作为输入变量,建立了GA-BP神经网络渗透率预测模型。将渗透率模型的预测结果与实测结果比较,两者具有较高的吻合度,证明该方法在煤储层参数预测的实践中具有较好的适应性。基于所建立的数学模型,对研究区的渗透率进行了预测,完成了渗透率平面分布图,为柳林地区煤层气的勘探开发提供了依据。

After analyzing the problems in the coal reservoir permeability prediction, the method predicting the permeability of coal reservoir by GA-BP neural network based on the logging information is put forward. Its mechanism and logging parameters standardized approach are analyzed. Six logging parameters are selected as input variables via grey correlation analysis, and the GA-BP neural network model to predict the permeability of coal reservoir based on well testing and log data from 56 wells in Liulin area is established. The permeability is predicted by using the model, and the prediction results are compared with the measured results. The two kinds of results have high agreement. The comparison results prove that the method of reservoir parameter prediction in coal has good applicability in practice. The permeability of the whole area is predicted based on the established mathematical model. The plane distribution map is completed, and it could provide basis for exploration and development of CBM in Liulin area.