Fracturing sand plugging risk early warning method based on deep learning and double logarithmic curves

• Main Authors: WANG LIN, ZHU NAN, WEI MIAN, LIU YIJIA, HUANG QI, WANG FANGXIANG, GENG TENG, YANG XIN, DONG BINGXIANG, LI NAN, LIN CHUNLAI, DU HUIMIN, SHAO JUNLONG
• Format: Patent
• Language: Chinese; English
• Published: 11.06.2024
• Subjects: CALCULATING; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention discloses a fracturing sand plugging risk early warning method based on deep learning and a double logarithmic curve. The method comprises the following steps that S1, multiple construction parameter time sequences in the fracturing construction process are collected; s2, on the basis of the time sequence in the step S1, predicting a future time sequence of each construction parameter by using a time sequence time domain analysis algorithm; s3, constructing and training an L-Aresnet network so as to correct the predicted value of each construction parameter in the future time sequence; s4, according to the time sequence change curve of the future sample value of each construction parameter obtained in the step S3, calculating a log value lgK of the slope K of the future change curve of each construction parameter as an early warning judgment value, and comparing the log value lgK with an early warning critical value; according to the method, multiple construction parameters, easy to monitor, of