BP模型在南海神狐海域天然气水合物储量参数预测中的应用

摘要

The BP neural network method has been used to forecast gas hydrates reservoir parameters ( porosity and saturation) based on the previous method by using logging data to forecast reservoir. Taken an measured well as an example, a network model is built with the well's logging data. The authors inputted the logging data of other wells and obtain the results of reservoir parameters, which are more accurate than the empirical formula through practice tests.%在用测井数据预测储量参数方法的基础上,采用BP神经网络法预测天然气水合物储量参数(孔隙度、饱和度).选取一口有实测值的井,将其测井数据作为样本数据,建立网络模型,由其他井的测井数据输入此模型得到储量参数预测结果.经过实践检验此模型得出的结果比经验公式法更精确.