Shale pressure fracture network prediction method based on deep learning
Fracturing evaluation is an important part of hydraulic fracturing,in which the acquisition of the fracture network param-eters after fracturing is the key section.In view of the problems such as high cost,operating difficulty and vulnerability of the com-monly used fracture diagnosis methods(microseismic monitoring technology,clinometer crack monitoring technology and distributed optical fiber monitoring technology),the target parameters herein were obtained based on the deep learning algorithm with significant advantages in nonlinear regression.The process was started with the construction fracturing curve,and comprehensively considered various theoretical models and empirical equations.For the quantization of fracturing curve,relevant parameters were selected as input,and the fracture network parameters(length,width and height of fracture network)of microseismic monitoring were taken as target parameters,and the relationship between them and target parameters was constructed by deep learning algorithm(BP neural network).The results show that the method has high prediction accuracy,and the average relative error is less than 12%.In addition,it is also found that the results are affected by the type and number of input parameters,network structure,and hyperparameters.
NETL
NSTL
万方数据
