Search

On the basis of the National Research University "MPEI", within the framework of the project to create a busbar with polymer-gas insulation for the transmission of electricity from offshore wind turbines, a study is being carried out to develop insulating structures for high-voltage busbars. The work is devoted to the development and justification of the design of a special conductor with combined insulation (polymer-gas) for laying in sea water for the transmission of electrical energy from wind turbines, where a polymer coating of current-carrying elements is used to increase the electrical strength and reduce the dimensions of the current conductor, and for to reduce electricity losses, the main gas insulation is used. Technical solutions for the execution of linear and connecting sections of bus ducts with polymer-gas insulation are presented. An analysis of the distribution of the thermal field in the coaxial gap was performed, demonstrating the features of design solutions for linear sections. According to the calculations, conclusions were drawn about the possibility of using combined insulation in special conductors.

We consider distributed optimization over several devices, each sending incremental model updates to a central server. This setting is considered, for instance, in federated learning. Various schemes have been designed to compress the model updates in order to reduce the overall communication cost. However, existing methods suffer from a significant slowdown due to additional variance omega > 0 coming from the compression operator and as a result, only converge sublinearly. What is needed is a variance reduction technique for taming the variance introduced by compression. We propose the first methods that achieve linear convergence for arbitrary compression operators. For strongly convex functions with condition number kappa, distributed among n machines with a finite-sum structure, each worker having less than in components, we also (i) give analysis for the weakly convex and the non-convex cases and (ii) verify in experiments that our novel variance reduced schemes are more efficient than the baselines. Moreover, we show theoretically that as the number of devices increases, higher compression levels are possible without this affecting the overall number of communications in comparison with methods that do not perform any compression. This leads to a significant reduction in communication cost. Our general analysis allows to pick the most suitable compression for each problem, finding the rig ht balance between additional variance and communication savings. Finally, we also (iii) give analysis for arbitrary quantized updates.

High-entropy alloys are promising materials for novel thin-film resistors since they have high resistivity and a low-temperature coefficient of resistivity (TCR). In this work, a new high-entropy thin-film CoCrFeNiTix was deposited on a Si/SiO2 substrate by means of magnetron sputtering of the multi-component target produced by hot pressing of the powder mixture. The samples possessed a thickness of 130–230 nm and an amorphous atomic structure with nanocrystallite traces. This structure persisted after being annealed up to 400 °C, which was confirmed using X-ray and electron diffraction. The film had a single-phase structure with a smooth surface and a uniform distribution of all elements. The obtained film served for microresistor elaboration, which was produced using the lithography technique and tested in a temperature range from −60 °C up to 200 °C. Resistivity at room temperature was estimated as 2.37 μOhm·m. The results have demonstrated that TCR depends on temperature according to the simple linear law in a range from −60 °C up to 130 °C, changing its value from −78 ppm/°C at low temperatures to −6.6 ppm/°C at 130 °C. Such characteristics show the possibility of using these high-entropy alloy films for resistive elements in contemporary and future micro-electronic devices.

The ordered Ca3TaGa3Si2O14 and disordered La3Ga5SiO14 crystals of the lantangallium silicate family were grown by the Czochralski method. The independent coefficients of thermal expansion of crystals αc and αa were determined by X-ray powder diffraction based on the analysis of X-ray diffraction spectra measured in the temperature range of 25÷1000°C. It is shown that in the temperature range of 25÷800°C the thermal expansion coefficients are linear. At temperatures above 800°C there is a nonlinear character of the thermal expansion coefficients associated with a decrease in the Ga content in the crystal lattice of the lantangallium silicate family crystals.

With increasing rates of anxiety and mood disorders across the world, there is an unprecedented need for preclinical animal models to generate translational results for humans experiencing disruptive emotional symptoms. Considering that life events resulting in a perception of loss are correlated with depressive symptoms, the enrichment-loss rodent model offers promise as a translational model for stress-initiated psychiatric disorders. Additionally, predisposed temperament characteristics such as coping styles have been found to influence an individual's stress response. Accordingly, male rats were profiled as either consistent or flexible copers and assigned to one of three environments: standard laboratory housing, enriched environment, or enriched environment exposure followed by downsizing to standard laboratory cages (i.e., enrichment-loss group). Throughout the study, several behaviors were assessed to determine stress, social, and reward-processing responses. To assess recovery of the stress response, fecal samples were collected following the swim stress in 3-h increments to determine the recovery trajectory of corticosterone (CORT) and dehydroepiandrosterone (DHEA) metabolite levels. Upon death, neural markers of neuroplasticity including doublecortin, glial fibrillary acidic factor, and brain-derived neurotrophic factor were assessed via immunohistochemistry. Results indicated the flexible coping animals in the continuous enriched group had higher DHEA/CORT ratios (consistent with adaptive responses in past research); furthermore, the enrichment-loss animals exhibited a blunted CORT response throughout the assessments and enriched flexible copers had faster CORT recovery rates than consistent copers. Standard housed animals exhibited less exploratory behavior in the open field task and continuous enriched, flexible rats consumed more food rewards than the other groups. No differences in neuroplasticity neural markers were observed. In sum, the results of the present study support past research indicating the disruptive consequences of enrichment-loss, providing evidence that the model represents a valuable approach for the investigation of neurobiological mechanisms contributing to interindividual variability in responses to changing experiential landscapes.

Determining and predicting reservoir formation properties for newly drilled wells represents a significant challenge for oil and gas companies. Extensive well logs are available only while or after drilling, and thus they bear substantial financial, technical, and operational risks. We propose a new machine learning data-based model for determining well properties similarity and further derive and predict well logs before drilling in a specific geological context. Our model starts with selecting crucial well intervals and aggregation of vital features that determine the petrophysical properties related to particular well layers. Then, a machine-learning algorithm uses this info as input to provide a similarity score between wells. Our fast-to-train nonlinear data-based model is a variant of gradient boosting. We show that this approach can work well in complex scenarios with missing data and inconsistent similarity measures. We compare the modern machine learning algorithms for the evaluation of well similarity models based on aggregated features. The algorithms include gradient boosting and baseline logistic regression models. Our assessment for a real well log dataset via group cross-validation demonstrates that the gradient boosting model pretty accurately identifies well similarity. The receiver operating characteristic quality metric (ROC AUC) is 0.824. The developed similarity learning framework provides a data-driven approach towards estimating well logs for planned and newly drilled wells. Therefore, it allows prediction, improves determination, and can drive an optimal selection of log measurements to be executed in a new well in a specific field / geological context.

Well log analysis, through deploying advanced artificial intelligence (AI) algorithms, is key for wellbore geological studies. By analyzing different well characteristics with modern AI tools it becomes possible to estimate interwell saturation with improved accuracy, outlining primary fluid channels and saturation propagations in the reservoirs interwell region. The development of modern deep learning and artificial intelligence methods allows analysts to predict interwell saturation as a function of observed data in the near wellbore logged geological layers. This work addresses the use of deep neural network architectures as well as tensor regression models for predicting interwell saturation from other well characteristics, such as resistivity and porosity, as well as local near-well saturation. Several algorithms are compared in terms of both accuracy and computational efficiency. Sensitivity analysis for model parameters is carried out, which is based on the wells’ geometry, radius, and multiple sampling techniques. Additionally, the impact of local saturation prior knowledge on the model accuracy is analyzed. A reservoir box model encompassing volumetric interwell porosity, resistivity and saturation data was utilized for the validating and testing of the AI algorithms. A prototype is developed with Python 3.6 programming language.

Combining physics-based models for well log analysis with artificial intelligence (AI) advanced algorithms is crucial for wellbore studies. Data-driven methods do not generalize well and lack theoretical knowledge accumulated in the field. Estimating well saturation significantly improves if predictions from physical models are used to constrain data-driven algorithms in outlined primary fluid channels and other important points of interest. Saturation propagations in the reservoirs interwell region also generalize better under using combination of models. This work addresses combined usage of theoretical and data-driven models by aggregating them into single hybrid model. Multiple physical and data-driven models are under study, their parameters are optimized using observations. Weighted sum is used to predict water saturation at every point with weights being recomputed at each step. Model outputs are compared in terms accuracy and cumulative loss. A synthesized reservoir box model encompassing volumetric interwell porosity, resistivity and saturation data is used for the validation of the algorithms. Aggregated model for estimating interwell saturation shows improved prediction accuracy compared both to physics-based or data-driven approaches separately.