A Multi-Level Non-Linear Solver for Complex Well Modelling

Complex well model has proved to be important for capturing the full physics in wellbore, including pressure losses, multiphase effects, and advanced device modelling. Numerical instability may be observed especially when the well is produced at a low rate from a highly productive multi-phase zone. In this paper, a new multi-level nonlinear solver is presented in a state-of-the-art parallel complex wellbore model for addressing some difficult numerical convergence problems. A sequential two-level nonlinear solver is implemented, where the inner solver is used to address the convergence in the constraint rate equation, and then the entire complex network is solved using an outer solver. Finally, the wellbore model is coupled with the grid solution explicitly, sequentially, or implicitly. This novel formulation is robust enough to greatly improve the numerical stability due to the lagging in the computation of mixture density in wellbore constraint rate equation and the variation in the fluid composition over Newton iterations in network nonlinear solver. The numerical challenge in the complex well model and the improvement of performance with the new nonlinear solver are demonstrated using reservoir simulation. Models with complex wells running into convergence problems are constructed and simulated. With this novel nonlinear solver, simulation gives much more reliable results on well productions without numerical oscillations and computational cost is much less.

Healthcare Data Analysis Using Water Wave Optimization-Based Diagnostic Model

This paper presents a new diagnostic model for various diseases. In the proposed diagnostic model, a water wave optimization (WWO) algorithm was implemented for improving the diagnosis accuracy. It was observed that the WWO algorithm suffered from the absence of global best information and premature convergence problems. Therefore in this work, some improvements were proposed to formulate the WWO algorithm as more promising and efficient. The global best information issue was addressed by using an improved solution search equation and the aim of this was to explore the global best optimal solution. Furthermore, a premature convergence problem was rectified by using a decay operator. These improvements were incorporated in the propagation and refraction phases of the WWO algorithm. The proposed algorithm was integrated into a diagnostic model for the analysis of healthcare data. The proposed algorithm aimed to improve the diagnosis accuracy of various diseases. The diverse disease datasets were considered for implementing the performance of the proposed diagnostic model based on accuracy and F-score performance indicators, while the existing techniques were regarded to compare the simulation results. The results confirmed that the WWO-based diagnostic model achieved a higher accuracy rate as compared to existing models/techniques with most disease/healthcare datasets. Therefore, it stated that the proposed diagnostic model is more promising and efficient for the diagnosis of different diseases.

On Ćirić-Prešić Operators in Metric Spaces

We show that the Prešić type operators of several variables can be regarded as an operator of a single variable and the fixed point problem of Prešić type can be regarded as a classical fixed point problem. We extend the recent result of Ćirić and Prešić by using the classical approach of Prešić. The key of the proof is based on the mappings introduced by Kada, Suzuki, and Takahashi. We also discuss the convergence problems of recursive real sequences and the Volterra integral equations as an application of our result.

A three-dimensional-model inversion of methyl chloroform to constrain the atmospheric oxidative capacity

Variations in the atmospheric oxidative capacity, largely determined by variations in the hydroxyl radical (OH), form a key uncertainty in many greenhouse and other pollutant budgets, such as that of methane (CH4). Methyl chloroform (MCF) is an often-adopted tracer to indirectly put observational constraints on large-scale variations in OH. We investigated the budget of MCF in a 4DVAR inversion using the atmospheric transport model TM5, for the period 1998–2018, with the objective to derive information on large-scale, interannual variations in atmospheric OH concentrations. While our main inversion did not fully converge, we did derive interannual variations in the global oxidation of MCF that bring simulated mole fractions of MCF within 1 %–2 % of the assimilated observations from the NOAA-GMD surface network at most sites. Additionally, the posterior simulations better reproduce aircraft observations used for independent validation compared to the prior simulations. The derived OH variations showed robustness with respect to the prior MCF emissions and the prior OH distribution over the 1998 to 2008 period. Although we find a rapid 8 % increase in global mean OH concentrations between 2010 and 2012 that quickly declines afterwards, the derived interannual variations were typically small (< 3 %/yr), with no significant long-term trend in global mean OH concentrations. The inverse system found strong adjustments to the latitudinal distribution of OH, relative to widely used prior distributions, with systematic increases in tropical and decreases in extra-tropical OH concentrations (both up to 30 %). These spatial adjustments were driven by intrahemispheric biases in simulated MCF mole fractions, which have not been identified in previous studies. Given the large amplitude of these adjustments, which exceeds spread between literature estimates, and a residual bias in the MCF intrahemispheric gradients, we suggest a reversal in the extratropical ocean sink of MCF in response to declining atmospheric MCF abundance (as hypothesized in Wennberg et al., 2004). This ocean source provides a more realistic explanation for the biases, possibly complementary to adjustments in the OH distribution. We identified significant added value in the use of a 3D transport model, since it implicitly accounts for variable transport and optimizes the observed spatial gradients of MCF, which is not possible in simpler models. However, we also found a trade-off in computational expense and convergence problems. Despite these convergence problems, the derived OH variations do result in an improved match with MCF observations relative to an interannually repeating prior for OH. Therefore, we consider that variations in OH derived from MCF inversions with 3D models can add value to budget studies of long-lived gases like CH4.

Margin CosReid Network for Pedestrian Re-Identification

This paper proposes a margin CosReid network for effective pedestrian re-identification. Aiming to overcome the overfitting, gradient explosion, and loss function non-convergence problems caused by traditional CNNs, the proposed GBNeck model can realize a faster, stronger generalization, and more discriminative feature extraction task. Furthermore, to enhance the classification ability of the softmax loss function within classes, the margin cosine softmax loss (MCSL) is proposed through a boundary margin introduction to ensure intraclass compactness and interclass separability of the learning depth features and thus to build a stronger metric-based learning model for pedestrian re-identification. The effectiveness of the margin CosReid network was verified on the mainstream datasets Market-1501 and DukeMTMC-reID compared with other state-of-the-art pedestrian re-identification methods.

Новые результаты в генеалогической классификации тюркских диалектов («случаи с аффрикатами»)

Introduction. As is well known, the three Turkic dialectal continua — Tatar-Bashkir, Shor-Khakass-Chulym, and Karachay-Balkar ones — have developed quite distinctive reflexes of proto-Turkic palatal *j- and *č-, *-č(-). While compiling the Dialectological Atlas of Russia’s Turkic Languages, the authors were able to compose exact isoglosses of *j- and *č change in members of the mentioned continua, which made it also possible to partially reevaluate genetic clusterization on the basis of this data. Materials and Methods. Apart from the available publications and archival sources on the three areas in question, the analysis is based on the authors’ extensive field work that involves the use of a set of lexical questionnaires compiled in accordance with known aspects of the Turkic linguistic history. The source recordings for every speaker were turned into idiolectal audio-dictionaries and are linked to an electronic etymological database of the Turkic languages, each elicitation analyzed both with the comprehension method and the software for experimental phonetics. Results. As it turns out, this methodology of field work and post-analysis provides information crucial for detailed linguistic clusterization of dialectal continua in particular and any dialectal system in general. Traditionally, subtle problems of divergence and convergence, problems of archaic and innovative phenomena receive their solutions. The results are as follows. Palatal *j- and *č in the languages of the Khakass-Shor-Chulym group have changed by a strict series of rules none of which could be simultaneous, nor could follow each other in a different order. Thus, the two Middle Chulym dialects — Melet and Tutal ones — prove to lack an immediate linguistic ancestor, the Tutal ‘dialect’ is an archaic version of Mrassu Shor, while Melet is closely related to Kyzyl Khakass. Reflexes of *j- and *č are also principal isoglosses for a previously undocumented Khakass dialect, which does not have any specific affinity with Saghai, Kyzyl and Kachin dialects. Areal analysis of KarachayBalkar shows that dz < proto-Turkic *j- is a secondary development, while, on the other hand, it is finally proven that reflexes *j- > dz~dʑ and *j- > ʑ~z form a more significant isogloss. And for the Tatar-Bashkir dialectal continuum, there were identified three main types of proto-Turkic *jreflexation; a chronology for these three types intermixing during the early period of the continuum is also proposed.

Design of Anchorage Zones of Pretensioned Concrete Girders: A Comparison of Nonlinear 3D FEM Results with Measurements on a Full Scale Beam

Pretensioned concrete beams are widely used for constructing large load-bearing structures and bridging long spans. Crack formation may occur in the end zones of these elements due to tensile splitting, spalling and bursting actions. Investigation of these zones is typically done by means of analytical methods, strut and tie modelling, 2D linear or nonlinear analysis, or full 3D nonlinear analysis. Especially challenging in this last approach is the modelling of the force transfer from the strands to the surrounding concrete as it strongly influences the magnitude of the tensile stresses. This paper presents a 3D nonlinear analysis of the anchorage zone of a pretensioned girder, and a comparison with experimental results (mechanical strain measurements, embedded strain gauges). Material modelling, steel-concrete interaction properties, as well as convergence problems are addressed systematically. The comparison indicates that a good agreement is found, both for concrete and rebar strains, and that a friction coefficient of 0.7 can be adopted, although the results for values from 0.5 to 0.9 do not differ that much. The results demonstrate that a 3D nonlinear analysis provides an excellent insight in the behavior of the end zones of pretensioned girders which opens perspectives for an optimization of the end zone design based on this type of analysis.

Graphene memristive synapses for high precision neuromorphic computing

Memristive crossbar architectures are evolving as powerful in-memory computing engines for artificial neural networks. However, the limited number of non-volatile conductance states offered by state-of-the-art memristors is a concern for their hardware implementation since trained weights must be rounded to the nearest conductance states, introducing error which can significantly limit inference accuracy. Moreover, the incapability of precise weight updates can lead to convergence problems and slowdown of on-chip training. In this article, we circumvent these challenges by introducing graphene-based multi-level (>16) and non-volatile memristive synapses with arbitrarily programmable conductance states. We also show desirable retention and programming endurance. Finally, we demonstrate that graphene memristors enable weight assignment based on k-means clustering, which offers greater computing accuracy when compared with uniform weight quantization for vector matrix multiplication, an essential component for any artificial neural network.

Pressure-Driven Simulation of Water Distribution Networks: Searching for Numerical Stability

EPANET uses a demand-driven approach to compute pressures and flows in the water distribution system. The demand-driven approach (DDA) assumes that the required demand is always fully satisfied no matter the existing pressure. In scenarios of pressure-deficient conditions the DDA results are not accurate, and a pressure-driven approach (PDA) is needed. Frequently, the PDA is accomplished by using equations that compute the available demand/leakage as a function of the current pressure. However, embedding such equations into the solver introduces convergence problems. This paper details the actions taken in WaterNetGen—an EPANET extension—to bring numerical stability to the pressure-driven solver, namely, by smoothing the pressure–demand/leakage relationship and the pump curve.

Improved SiC MOSFET SPICE Model to Avoid Convergence Errors

This paper presents improvements to a SPICE model for a commercially available SiC MOSFET to avoid convergence errors while still providing reliable simulation results. Functionality in the internal part of the model that shapes the transconductance of the device according to its junction temperature and gate-source voltage dependency has been improved to provide a continuous characteristic rather than the initial discontinuous performance. Furthermore, the output characteristics from the initial and the proposed model have been compared to lab measurements of an actual device. The results show that the proposed and initial model provide equally reliable simulation results. However, the proposed model does not run into convergence problems.