Design And Implementation Of Logging Data Visualization And Reconstruction System

There is a wealth of information in well logging data and the display of logging curves can be of great practical importance in providing oilfield researchers with intuitive and reliable information on downhole parameters.However,in the actual logging process,due to machine failure or manual operation errors,the logging data sometimes has the problem of missing continuity,so it is valuable to predict and repair the missing logging curves.Therefore,this thesis proposes a hybrid model to realize logging data reconstruction,and designs and develops a logging data visualization and reconstruction system to realize visual display of logging data and related data management functions while hooking up the algorithm model proposed in this thesis into the system,so that operators can complete the reconstruction of missing data by means of software operation,realizing the transformation of deep learning theoretical analysis to practical application.The details of the research are as follows.(1)To address the problem of missing logging curves,this thesis constructs a hybrid CNN-GRU model to reconstruct missing logging curves by investigating the principles of Convolutional Neural Network(CNN)and Gated Recurrent Unit(GRU)models.Using the feature extraction capability of CNN,the main features were extracted from the original logging data and then input into the GRU model to fit the sequence data.Through the analysis of the prediction experimental results and the comparison with the results of deep neural network,recurrent neural network,long and short-term memory neural network and GRU neural network models,it is proved that the CNN-GRU hybrid model constructed in this thesis has higher prediction accuracy and has some practical significance in the reconstruction of logging data.(2)Conduct a requirement analysis of the logging data visualisation and reconstruction system,select the system development model and design the overall functional structure of the system.Based on the.NET system development framework and C# language,and using Microsoft Visual Studio 2019 system development software and My SQL database tools,the logging data visualization and reconstruction system of this topic is developed.The Pyinstaller was used to encapsulate the model designed in this thesis into an executable file and execute it using the Process class in C#,solving the problem of compatibility between Python and C#.The final result is a visual display of logging data and reconstruction of missing data,as well as data storage,historical data playback,data export,log management and other related functions.The system designed in this thesis has a user-friendly interface,high stability,rich functionality and can invoke the algorithm model built in this thesis to achieve the curve reconstruction function,which provides a feasible application of deep learning based logging curve reconstruction technology to practical engineering.