Wellbore Strengthening Design Criteria and Enhanced Fracture Gradient for Widening Stable Mud Weight Window and Enabling Safe and Efficient Drilling in Depleted Reservoirs

Drilling in permeable formations, especially depleted reservoirs, can particularly benefit from simultaneous wellbore shielding and strengthening functionalities of drilling mud compounds. The ability to generate simultaneous wellbore shielding and strengthening in reservoirs has potential to widen stable mud weight windows to drill such reservoirs without the need to switch from wellbore strengthening compound to wellbore shielding compound, and vice-versa. Wellbore shielding and strengthening experiments were conducted on three outcrop sandstones with three mud compounds. The wellbore shielding stage was conducted by increasing the confining and borehole pressures in 4-5 steps until both reached target pressures. CT scan images demonstrate consistency of the filtration rates with observed CT scanned mud cakes which are dependent on the sandstone pore size and mud compound particle size distributions. In wellbore strengthening stage, the borehole pressure was increased until fracture was initiated, which was detected via borehole pressure trend and CT scan imaging. The fractures generated were observed to be plugged by mud filter solids which are visible in the CT scan images. The extent of observed fracture solid plugging varies with rock elastic properties, fracture width and mud compound particle size distribution. Based on the laboratory test data, fracture gradient enhancement concept was developed for the mud compounds. In addition, the data obtained and observations from the tests were used to develop optimal empirical design criteria and guidelines to achieve dual wellbore strengthening and shielding performance of the mud compounds. The design criteria were validated on a well which was treated with one of the mud compounds based on its mud loss events during drilling and running casing.

English Pronunciation Standards Based on Multimodal Acoustic Sensors

In recent years, economic globalization is the trend, and communication between countries is getting closer and closer; more and more people begin to pay attention to learning spoken English. The development of computer-aided language learning makes it more convenient for people to learn spoken English; however, the detection and correction of incorrect English pronunciation, which is its core, are still inadequate. In this paper, we propose a multimodal end-to-end English pronunciation error detection and correction model based on audio and video, which does not require phoneme forced alignment of the English pronunciation video signal to be processed, and uses rich audio and video features for English pronunciation error detection, which improves the error detection accuracy to a great extent especially in noisy environments. To address the shortcomings of the current lip feature extraction algorithm which is too complicated and not enough characterization ability, a feature extraction scheme based on the lip opening and closing angle is proposed. The lip syllable frames are obtained by video frame splitting, the syllables are denoised, the key point information of the lips is obtained using a gradient enhancement-based regression tree algorithm, the effects of speaker tilt and movement are removed by scale normalization, and finally, the lip opening and closing angles are calculated using mathematical geometry, and the lip feature values are generated by combining the angle changes.

Cyberattacks Detection and Analysis in a Network Log System Using XGBoost with ELK Stack

The usage of artificial intelligence and machine learning methods on cyberattacks increasing significantly recently. For the defense method of cyberattacks, it is possible to detect and identify the attack event by observing the log data and analyzing whether it has abnormal behavior or not. This paper implemented the ELK Stack network log system (NetFlow Log) to visually analyze log data and present several network attack behavior characteristics for further analysis. Additionally, this system evaluated the extreme gradient enhancement (XGBoost), Recurrent Neural Network (RNN), and Deep Neural Network (DNN) model for machine learning methods. Keras was used as a deep learning framework for building a model to detect the attack event. From the experiments, it can be confirmed that the XGBoost model has an accuracy rate of 96.01% for potential threats. The full attack data set can achieve 96.26% accuracy, which is better than RNN and DNN models.

SOLVING THE PROBLEM OF OIL RELEASE USING NEURAN NETWORKS

The problem of oil displacement was solved through neural networks. The Buckley-Leverett model was chosen, which describes the process of displacing oil with water. It consists of the equation of continuity of oil and water phases and Darcy's law. The task is to optimize the problem of oil displacement. Optimization is carried out at three levels: vectorization of calculations; implementation of the algorithm using neural networks. The peculiarity of the method proposed in the work is the identification of the method with high accuracy and minimal errors, the solution with the help of neural networks. The study is also one of the first to compare neural and recurrent neural networks. As a result of the study, gradient enhancement classifiers and neural networks showed high accuracy, 99.99% and 97.4%, respectively. To achieve this goal, more than 67,000 data sets from 10th grade were created. These data are important for solving the problem of oil displacement in porous media. The proposed method provides a simple and sophisticated way to introduce oil knowledge into neural networks. This eliminates two of the most important disadvantages of neural networks: the need for large data sets and the reliability of extrapolation. The proposed principles can be summarized in countless ways in the future and should lead to the creation of a new class of algorithms for solving direct and reverse oil problems.

On the Formation of Tropopause Folds and Constituent Gradient Enhancement Near Westerly Jets

The role of differential advection in creating tropopause folds and strong constituent gradients near midlatitude westerly jets is investigated using the University of Wisconsin Non-hydrostatic Modeling System (UWNMS). Dynamical structures are compared with aircraft observations through a fold and subpolar jet (SPJ) during RF04 of the Stratosphere-Troposphere Analyses of Regional Transport (START08) campaign. The observed distribution of water vapor and ozone during RF04 provides evidence of rapid transport in the SPJ, enhancing constituent gradients above relative to below the intrusion. The creation of a tropopause fold by quasi-isentropic differential advection on the upstream side of the trough is described. This fold was created by a southward jet streak in the SPJ, where upper tropospheric air displaced the tropopause eastward in the 6-10 km layer, thereby overlying stratospheric air in the 3-6 km layer. The subsequent superposition of the subtropical and subpolar jets is also shown to result from quasi-isentropic differential advection.The occurrence of low values of ozone, water vapor, and potential vorticity on the equatorward side of the SPJ can be explained by convective transport of low-ozone air from the boundary layer, dehydration in the updraft, and detrainment of inertially-unstable air in the outflow layer. An example of rapid juxtaposition with stratospheric air in the jet core is shown for RF01. The net effect of upstream convective events is suggested as a fundamental cause of the strong constituent gradients observed in midlatitude jets. Idealized diagrams illustrate the role of differential advection in creating tropopause folds and constituent gradient enhancement.

APPLICATION OF MACHINE LEARNING ALGORITHMS FOR PROCESSING COMMENTS FROM THE YOUTUBE VIDEO HOSTING UNDER TRAINING VIDEOS

Purpose. Detecting toxic comments on YouTube video hosting under training videos by classifying unstructured text using a combination of machine learning methods. Methodology. To work with the specified type of data, machine learning methods were used for cleaning, normalizing, and presenting textual data in a form acceptable for processing on a computer. Directly to classify comments as “toxic”, we used a logistic regression classifier, a linear support vector classification method without and with a learning method – stochastic gradient descent, a random forest classifier and a gradient enhancement classifier. In order to assess the work of the classifiers, the methods of calculating the matrix of errors, accuracy, completeness and F-measure were used. For a more generalized assessment, a cross-validation method was used. Python programming language. Findings. Based on the assessment indicators, the most optimal methods were selected – support vector machine (Linear SVM), without and with the training method using stochastic gradient descent. The described technologies can be used to analyze the textual comments under any training videos to detect toxic reviews. Also, the approach can be useful for identifying unwanted or even aggressive information on social networks or services where reviews are provided. Originality. It consists in a combination of methods for preprocessing a specific type of text, taking into account such features as the possibility of having a timecode, emoji, links, and the like, as well as in the adaptation of classification methods of machine learning for the analysis of Russian-language comments. Practical value. It is about optimizing (simplification) the comment analysis process. The need for this processing is due to the growing volumes of text data, especially in the field of education through quarantine conditions and the transition to distance learning. The volume of educational Internet content already needs to automate the processing and analysis of feedback, over time this need will only grow.

A Finite Plasticity Gradient-Damage Model for Sheet Metals during Forming and Clinching

In recent years, clinching has gathered popularity to join sheets of different materials in industrial applications. The manufacturing process has some advantages, as reduced joining time, reduced costs, and the joints show good fatigue properties. To ensure the joint strength, reliable simulations of the material behaviour accounting for process-induced damage are expected to be beneficial to obtain credible values for the ultimate joint strength and its fatigue limit. A finite plasticity gradient-damage material model is outlined to describe the plastic and damage evolutions during the forming of sheet metals, later applied to clinching. The utilised gradient-enhancement cures the damage-induced localisation by introducing a global damage variable as an additional finite element field. Both, plasticity and damage are strongly coupled, but can, due to a dual-surface approach, evolve independently. The ability of the material model to predict damage in strongly deformed sheets, its flexibility and its regularization properties are illustrated by numerical examples.

Comparison of evolving gradient damage formulations with different activity functions

In the paper, two existing upgrades of the gradient damage model for the simulations of cracking in concrete are compared. The damage theory is made nonlocal via a gradient enhancement to overcome the mesh dependence of simulation results. The implicit gradient model with an averaging equation, where the internal length parameter is assumed as constant during the strain softening analysis, gives unrealistically broadened damage zones. The gradient enhancement of the scalar damage model can be improved via a function of an internal length scale, so an evolution of the gradient activity is postulated during the localization process. Two different modifications of the averaging equation and respective evolving gradient damage formulations are presented. Different activity functions are tested to see whether the formation of a too wide damage zone still occurs. Activating or localizing character of the gradient influence can be introduced and the impact of both approaches on the numerical results is shown in the paper. The aforementioned variants are implemented and examined using the benchmarks of tension in a bar and bending of a cantilever beam.

Size effect analysis of quasi-brittle fracture with localizing gradient damage model

The size effect of a quasi-brittle fracture is associated with the size of fracture process zone relative to the structural characteristic length. In numerical simulations using damage models, the nonlocal enhancement is commonly adopted to regularize the softening response. However, the conventional nonlocal enhancement, both integral and gradient approaches, induces a spurious spreading of damage zone. Since the evolution of fracture process zone cannot be captured well, the conventional nonlocal enhancement cannot predict the size effect phenomenon accurately. In this paper, the localizing gradient enhancement is adopted to avoid the spurious spreading of damage. Considering the three-point bend test of concrete beams, it is demonstrated that the dissipation profiles obtained with the localizing gradient enhancement compare well with those of reference meso-scale lattice models. With the correct damage evolution process, the localizing gradient enhancement is shown to capture the size effect phenomenon accurately for a series of geometrically similar concrete beams, using only a single set of material parameters.