利用常规测井曲线及神经网络求取储层裂缝密度

Using Conventional Logs and BP Mode For Reservoir Fracture Density

裂缝性储层参数的研究是裂缝性油气藏勘探的一个重点。本文以裂缝密度为基础,利用不同小波基对常规测井中的分维后的密度测井数据分别共进行了10次分解并与裂缝密度进行相关性分析,进而建立合适的数据体,为准确提取裂缝信息并计算裂缝密度建立基础。分析结果表明,小波分解的d7信号与裂缝密度线性相关性最好,使用该数据建立数据体,基于神经网络方法,计算出的裂缝密度准确度较高。

The study of fractured reservoir parameters is a key part of fractured reservoir exploration. In this paper, based on fracture density, we calculate correlation of fractal dimensions data which has been attend from conventional density logs data transform after 10 times using different wavelets , and then establish the appropriate data volume for accurate information of fracture and establishing the basis of calculation of fracture density. The results show that linear correlation between d7 of wavelet decomposition signal and fracture density is the best, and using the data to establish data volume based on neural network, we can get fracture density accurately.