AUTOMATED SYSTEM FOR IDENTIFICATION AND PREDICTION OF COMPLICATIONS IN THE PROCESS OF CONSTRUCTION OF OIL AND GAS WELLS

摘要

FIELD: oil and gas industry.;SUBSTANCE: invention relates to systems for monitoring the construction of oil and gas wells and control of drilling operations, and is intended to identify and predict complications of the main types, such as absorption of drilling fluid, sticking (tightening) of drilling tools, gas and oil water showings during the construction of oil and gas wells. The technical problem to be solved by the proposed invention is to reduce the accident rate during the construction of oil and gas wells by increasing the accuracy and reliability of identifying and predicting the occurrence of complications in a wide range of predicted types of complications and existing restrictions on the composition and volume of initial data. This problem is solved by the fact that the automated system for identifying and predicting complications in the construction of oil and gas wells contains a module for collecting real-time data of geological and technological research from the construction site with an archived database of geological and technological research connected to it, a drilling simulator, a simulator database, module for preliminary processing of geological and technological research data, module for marking up geological and technological research data, marked and unmarked databases for geological and technological research, module for forming, training and validating anomaly detection model in geological research data, module for forming, training and validation models for predicting the values ​​of functions of indicators of the occurrence of complications, a module for predicting the occurrence of anomalies in the data of geological and technological studies, a module for predicting the values ​​of functions of indicators of the occurrence of complications, a formation module, training and validation of a recurrent neural network model for predicting the occurrence of complications, a module for evaluating the predicted values ​​of the probabilities of complications and a module for analyzing and generating warnings about the occurrence of complications and emergencies. ;EFFECT: expanded the input data space for predicting the occurrence of complications during well construction through the use of additional sources of information - unlabeled data sets and simulation data from the drilling simulator, as well as the use of auxiliary machine learning models to improve the accuracy and reliability of the classification neural network model.;1 cl, 1 dwg