A Real-time Prediction Method of Formation Fracture Length Based on Kernel SMO Algorithm

Abstract

The embodiment of the present application provides a method for real-time prediction of formation fracture length based on the kernel SMO algorithm, which belongs to the field of data mining and the field of drilling fluid plugging. The method includes: collecting the historical drilling data of the target block, the actual value of the historical drilling fracture length and the real-time drilling data, and performing data preprocessing; dividing the preprocessed historical drilling data into a training set and a testing set, and determining the core value in advance. The number of iterations M and the allowable error e of the SMO perceptron; use the parameters related to the crack length to generate parameter vectors, initialize all parameter vectors, and calculate the kernel matrix of the parameter vector; according to the parameter vector, kernel matrix, iteration times and allowable error To generate a new perceptron model, and test the new model through the data of the test set, if the test result does not meet the accuracy requirements, the model will be regenerated; use the new model that meets the accuracy requirements to predict real-time drilling data, so as to achieve formation fractures The effect of length on real-time predictions.

Description

technical field [0001] The invention relates to a method for real-time prediction of formation fracture length based on a nuclear SMO algorithm, which belongs to the field of data mining and the field of drilling fluid loss plugging. Background technique [0002] Lost circulation is one of the most serious complex situations that affect the safety of drilling operations. Lost circulation not only brings great difficulties to drilling engineering, but also seriously affects the development speed of oil and gas resources. Although my country's plugging technology is becoming more and more mature, the plugging effect is not ideal. Most oil wells have a plugging success rate of less than 30% after the occurrence of lost circulation, and the repeated leakage of oil wells is extremely frequent. Whether it is the low plugging success rate or the frequent repeated leakage, it is largely due to the fact that the fracture length cannot be judged in advance. The unknown fracture lengt...

AI Technical Summary

Problems solved by technology

[0003] At present, there are two main methods for measuring the length of fractures in my country: (1) Judging the length of underground fractures based on numerical simulation methods. This method is only applicable to some areas with uncomplicated geological structures. In complicated situations, the accuracy rate will be seriously inaccurate; (2) Using imaging logging technology to judge the fracture length, although this method has high accuracy and strong practicability, the imaging logging process is complicated and expensive to measure, and a long well section measurement It will cost hundreds of thousands or even millions of RMB, so it is difficult to popularize

Images