Prediction of Equivalent Liquid Permeability from Gas Permeability Measurements: A Recurrent Neural Network Approach

2021 ◽  
Author(s):  
Husam Alkinani ◽  
Abo Taleb Al-Hameedi ◽  
Shari Dunn-Norman
Keyword(s):  
Neural Network ◽  
Recurrent Neural Network ◽  
Measurement Errors ◽  
Mean Squared Error ◽  
Gas Permeability ◽  
Core Sample ◽  
Reservoir Properties ◽  
Core Samples ◽  
Liquid Permeability ◽  
Hidden Layer

Abstract One of the most vital reservoir properties is permeability. It is usually measured using core samples with two major measurement methods; using gas or using liquid. The purpose of this work is to use a data-driven recurrent neural network model to estimate the equivalent liquid permeability based on gas permeability. By using this model, the equivalent liquid permeability can be predicted for the permeability of core samples with rich clay minerals measured using gas (or any core sample that is measured using gas). This will give an alternative way to the currently used method (Klinkenberg method). Core sample data measurements of more than 500 cores were obtained from limestone formations. The data went through a processing step to eliminate any measurement errors. Then, the data were clustered into training, validation, and testing. After many iterations, a decision was made to have a network with four hidden layer and twenty neurons in each hidden layer, and four delays in the input and the output. The findings showed that the network had stopped training after nine epochs with a validation mean squared error (MSE) of 5.3. The model exhibited excellent performance during training, validation, and testing with an overall R2 of 0.91 which is excellent. These findings prove that the model can closely track the actual equivalent liquid permeability measurements using the gas permeability measurements data within a reasonable margin of error. With the rise of machine learning and other artificial intelligence (AI) methods as well as the potential application in the petroleum industry, these methodologies can revolutionize the industry and save time and money.

scholarly journals PREDIKSI PENGGUNAAN BANDWIDTH MENGGUNAKAN ELMAN RECURRENT NEURAL NETWORK

2016 ◽  
Vol 10 (2) ◽  
pp. 127-135
Author(s):  
Jefri Radjabaycolle ◽  
Reza Pulungan
Keyword(s):  
Neural Network ◽  
Recurrent Neural Network ◽  
Minimum Error ◽  
Target Error ◽  
Hidden Layer ◽  
Elman Recurrent Neural Network

Jaringan Syaraf Tiruan (JST) sering dipakai dalam menyelesaikan permasalahan tertentu seperti prediksi, klasifikasi, dan pengolahan data. Berdasarkan hal tersebut, dalam penelitian ini mencoba menerapkan JST untuk menangani permasalahan dalam prediksi penggunaan bandwidth. Sistem yang dikembangkan dapat digunakan untuk memprediksi pengunaan bandwidth dengan menerapkan Elman Recurrent Neural Network (ERNN). Struktur Elman dipilih karena dapat membuat iterasi jauh lebih cepat sehingga memudahkan proses konvergensi.. Vektor input yang digunakan menggunakan windows size. Hasil penelitian dengan menggunakan target error sebesar 0.001 menunjukkan nilai MSE terkecil yaitu pada windows size 11 dengan nilai 0.002833. Kemudian dengan menggunakan 13 neuron pada hidden layer diperoleh nilai error paling optimal (minimum error) sebesar 0.003725.

scholarly journals Kp forecasting with a recurrent neural network

2019 ◽  
Vol 9 ◽  
pp. A19 ◽  
Author(s):  
Ernest Scott Sexton ◽  
Katariina Nykyri ◽  
Xuanye Ma
Keyword(s):  
Neural Network ◽  
Solar Wind ◽  
Recurrent Neural Network ◽  
Historical Data ◽  
Time Step ◽  
Prediction Horizon ◽  
Sigma B ◽  
Input Layer ◽  
Solar Wind Parameters ◽  
Hidden Layer

In an effort to forecast the planetary Kp-index beyond the current 1-hour and 4-hour predictions, a recurrent neural network is trained on three decades of historical data from NASA’s Omni virtual observatory and forecasts Kp with a prediction horizon of up to 24 h. Using Matlab’s neural network toolbox, the multilayer perceptron model is trained on inputs comprised of Kp for a given time step as well as from different sets of the following six solar wind parameters, Bz, n, V, |B|, σB and $ {\sigma }_{{B}_z}$. The purpose of this study was to test which combination of the solar wind and Interplanetary Magnetic Field (IMF) parameters used for training gives the best performance as defined by correlation coefficient, C, between the predicted and actually measured Kp values and Root Mean Square Error (RMSE). The model consists of an input layer, a single nonlinear hidden layer with 28 neurons, and a linear output layer that predicts Kp up to 24 h in advance. For 24 h prediction, the network trained on Bz, n, V, |B|, σB performs the best giving C in the range from 0.8189 (for 31 predictions) to 0.8211 (for 9 months of predictions), with the smallest RMSE.

scholarly journals Sistem Evaluasi Kelayakan Mahasiswa MagangMenggunakan Elman Recurrent Neural Network

2014 ◽  
Vol 8 (1) ◽  
pp. 37 ◽  
Author(s):  
Agus Aan Jiwa Permana ◽  
Widodo Prijodiprodjo
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Recurrent Neural Network ◽  
Skills Training ◽  
Minimum Error ◽  
Training Mode ◽  
Artificial Neural ◽  
Artificial Neural Network Ann ◽  
Hidden Layer ◽  
Elman Recurrent Neural Network

AbstrakJaringan Syaraf Tiruan (JST) dapat digunakan untuk memecahkan permasalahan tertentu seperti prediksi, klasifikasi, pengolahan data, dan robotik.Berdasarkan paparan tersebut, sehingga dalam penelitian ini mencoba menerapkan JST untuk menangani permasalahan dalam program magang yang sedang dihadapi dalam upaya untuk meningkatkan kompetensi, pengalaman, serta melatih softskill mahasiswa.Sistem yang dikembangkan dapat digunakan untuk mengevaluasi kelayakan mahasiswa dalam program magang ke luar daerah dengan menerapkan Elman Recurrent Neural Network (ERNN), sehingga dapat memberikan informasi yang akurat kepada pihak jurusan untuk menentukan keputusan yang tepat.Struktur Elman dipilih karena dapat membuat iterasi jauh lebih cepat sehingga memudahkan proses konvergensi. Adapun metode pembelajaran yang digunakan adalah Backpropagation ThroughTime dengan model epochwise training mode. Sistem diimplementasikan dengan menggunakan bahasa pemrograman C# dengan basis data MySQL. Vektor input yang digunakan terdiri dari 11 variabel. Hasil penelitian menunjukkan bahwa sistem yang dikembangkan akan cepat mengalami konvergen dan mampu mencapai nilai error paling optimal (minimum error) apabila menggunakan 1 hidden layer dengan jumlah neuron 20 unit. Akurasi terbaik dapat diperoleh dengan menggunakan LR sebesar 0.01 dan momentum 0.85 dimana akurasi rata-rata dalam pengujian mencapai 87.50%. Kata kunci—Evaluasi, Kelayakan, Jaringan Syaraf Tiruan (JST), Elman Recurrent Neural Network, Magang Abstract Artificial Neural Network (ANN) can be used to solve specific problems such as prediction, classification, data processing, and robotics. Based on the exposure, so in this study tried to apply neural networks to handle problems in apprentice program facing in an effort to increase the competence, experience and soft skills training students. The system developed can be used to evaluate the students in the apprentice program to other regions by applying the Elman Recurrent Neural Network (ERNN), so it can provide accurate information to the department to determine appropriate decisions. Elman structure was chosen because it can be create much more rapidly iterations so as to facilitate the convergence process. The learning method used is Backpropagation Through Time with model epochwise training mode. The system is implemented using the C # programming language with a MySQL database. Input vector used consists of 11 variables. The results showed that the developed system will rapidly converge and can reach optimal error value (minimum error) when using one hidden layer with 20 units number of neurons. Best accuracy can be obtained using the LR of 0.01 and momentum 0.85 which average accuracy reaches 87.50% in testing. Keywords—Evaluation, Feasibility, Artificial Neural Network (ANN), Elman Recurrent Neural Network, Apprenticeship

Software Reliability Prediction using Fuzzy Min-Max Algorithm and Recurrent Neural Network Approach

2016 ◽  
Vol 6 (4) ◽  
pp. 1929 ◽  
Author(s):  
Manmath Kumar Bhuyan ◽  
Durga Prasad Mohapatra ◽  
Srinivas Sethi
Keyword(s):  
Neural Network ◽  
Recurrent Neural Network ◽  
Software Reliability ◽  
Reliability Prediction ◽  
Software Systems ◽  
Software Projects ◽  
Neural Network Approach ◽  
Data Set ◽  
Failure Data ◽  
Hidden Layer

Fuzzy Logic (FL) together with Recurrent Neural Network (RNN) is used to predict the software reliability. Fuzzy Min-Max algorithm is used to optimize the number of the kgaussian nodes in the hidden layer and delayed input neurons. The optimized recurrent<br />neural network is used to dynamically reconfigure in real-time as actual software failure. In this work, an enhanced fuzzy min-max algorithm together with recurrent neural network based machine learning technique is explored and a comparative analysis is performed for the modeling of reliability prediction in software systems. The model has been applied on data sets collected across several standard software projects during system testing phase with fault removal. The performance of our proposed approach has been tested using distributed system application failure data set.

Estimating normal moveout velocity using the recurrent neural network

2019 ◽  
Vol 7 (4) ◽  
pp. T819-T827
Author(s):  
Reetam Biswas ◽  
Anthony Vassiliou ◽  
Rodney Stromberg ◽  
Mrinal K. Sen
Keyword(s):  
Neural Network ◽  
Recurrent Neural Network ◽  
Mean Squared Error ◽  
Correction Term ◽  
Computation Time ◽  
Velocity Estimation ◽  
Data Set ◽  
Arrival Times ◽  
Nmo Correction ◽  
Novel Method

Machine learning (ML) has recently gained immense popularity because of its successful application in complex problems. It develops an abstract relation between the input and output. We have evaluated the application of ML to the most basic seismic processing of normal moveout (NMO) correction. The arrival times of reflection events in a common midpoint (CMP) gather follow a hyperbolic trajectory; thus, they require a correction term to flatten the CMP gather before stacking. This correction term depends on an rms velocity, also referred to as the NMO velocity. In general, NMO velocity is estimated using the semblance measures and picking the peaks in the velocity panel. This process requires a lot of human intervention and computation time. We have developed a novel method using one of the tools based on an ML- approach and applied to the NMO velocity estimation problem. We use the recurrent neural network (RNN) to estimate the NMO velocity directly from the seismic data. The input to the network is a seismic gather and corresponding precalculated NMO velocity (as prelabeled data set) to flatten the gather. We first train the network to develop a relationship between the input gathers (before NMO correction) and the corresponding NMO velocities for a few CMPs as a supervised learning process. Adam optimization algorithm is used to train the RNN. The output from the network is then compared against the correct NMO velocity. The error between the two velocities is then used to update the weight of the neurons and to minimize the mean-squared error between the two velocities. After the network is trained, it can be used to calculate the NMO velocity for the rest of the seismic gathers. We evaluate our method on a noisy data set from Poland. We used only 10% of the CMPs to train the network, and then we used the trained network to predict NMO velocity for the remaining CMP locations. The stack section obtained by using RNN-generated NMO velocities is nearly identical to that obtained by the conventional semblance method.

scholarly journals A fuzzy recurrent neural network for driver fatigue detection based on steering-wheel angle sensor data

2019 ◽  
Vol 15 (9) ◽  
pp. 155014771987245 ◽  
Author(s):  
Zuojin Li ◽  
Qing Yang ◽  
Shengfu Chen ◽  
Wei Zhou ◽  
Liukui Chen ◽  
...  
Keyword(s):  
Neural Network ◽  
Time Series ◽  
Recurrent Neural Network ◽  
Recognition Rate ◽  
Feature Learning ◽  
Sensor Data ◽  
Steering Wheel ◽  
Input Layer ◽  
Driving Conditions ◽  
Hidden Layer

The study of the robust fatigue feature learning method for the driver’s operational behavior is of great significance for improving the performance of the real-time detection system for driver’s fatigue state. Aiming at how to extract more abstract and deep features in the driver’s direction operation data in the robust feature learning, this article constructs a fuzzy recurrent neural network model, which includes input layer, fuzzy layer, hidden layer, and output layer. The steering-wheel direction sensing time series sends the time series to the input layer through a fixed time window. After the fuzzification process, it is sent to the hidden layer to share the weight of the hidden layer, realize the memorization of the fatigue feature, and improve the feature depth capability of the steering wheel angle time sequence. The experimental results show that the proposed model achieves an average recognition rate of 87.30% in the fatigue sample database of real vehicle conditions, which indicates that the model has strong robustness to different subjects under real driving conditions. The model proposed in this article has important theoretical and engineering significance for studying the prediction of fatigue driving under real driving conditions.

scholarly journals Intelligent Noise Removal from EMG Signal Using Focused Time-Lagged Recurrent Neural Network

2009 ◽  
Vol 2009 ◽  
pp. 1-12 ◽  
Author(s):  
S. N. Kale ◽  
S. V. Dudul
Keyword(s):  
Neural Network ◽  
Recurrent Neural Network ◽  
Biomedical Application ◽  
Noise Removal ◽  
Noise Signal ◽  
Modern Human ◽  
Emg Signal ◽  
Uniform Noise ◽  
Time Lagged ◽  
Hidden Layer

Electromyography (EMG) signals can be used for clinical/biomedical application and modern human computer interaction. EMG signals acquire noise while traveling through tissue, inherent noise in electronics equipment, ambient noise, and so forth. ANN approach is studied for reduction of noise in EMG signal. In this paper, it is shown that Focused Time-Lagged Recurrent Neural Network (FTLRNN) can elegantly solve to reduce the noise from EMG signal. After rigorous computer simulations, authors developed an optimal FTLRNN model, which removes the noise from the EMG signal. Results show that the proposed optimal FTLRNN model has an MSE (Mean Square Error) as low as 0.000067 and 0.000048, correlation coefficient as high as 0.99950 and 0.99939 for noise signal and EMG signal, respectively, when validated on the test dataset. It is also noticed that the output of the estimated FTLRNN model closely follows the real one. This network is indeed robust as EMG signal tolerates the noise variance from 0.1 to 0.4 for uniform noise and 0.30 for Gaussian noise. It is clear that the training of the network is independent of specific partitioning of dataset. It is seen that the performance of the proposed FTLRNN model clearly outperforms the best Multilayer perceptron (MLP) and Radial Basis Function NN (RBF) models. The simple NN model such as the FTLRNN with single-hidden layer can be employed to remove noise from EMG signal.

Sign in / Sign up

Export Citation Format

Share Document