Large System Sizes Research Articles

Inverse problems for structured datasets using parallel TAP equations and restricted Boltzmann machines

We propose an efficient algorithm to solve inverse problems in the presence of binary clustered datasets. We consider the paradigmatic Hopfield model in a teacher student scenario, where this situation is found in the retrieval phase. This problem has been widely analyzed through various methods such as mean-field approaches or the pseudo-likelihood optimization. Our approach is based on the estimation of the posterior using the Thouless–Anderson–Palmer (TAP) equations in a parallel updating scheme. Unlike other methods, it allows to retrieve the original patterns of the teacher dataset and thanks to the parallel update it can be applied to large system sizes. We tackle the same problem using a restricted Boltzmann machine (RBM) and discuss analogies and differences between our algorithm and RBM learning.

Quasi-Solid-State Electrolyte Using an In Situ Click Reaction for Safety-Enhanced Lithium-Ion Batteries

Lithium-ion batteries with high energy density have been used widely as electrochemical energy storage devices to power mobile electronics, electric vehicles and large-sized energy storage systems. However, there are still safety concerns related to the highly flammable liquid electrolytes. As a highly safe electrolyte, we synthesized a three-dimensional quasi-solid-state electrolyte with high ionic conductivity of 2.1 × 10−3 S cm−1 using an in situ click reaction. Perfluoroether diacrylate with high oxidative stability was synthesized and used in in situ click reaction along with pentaerylythritol tetrakis(3-mercaptoptopionate) as the cross-linking agents. The lithium-ion cell composed of a graphite anode, a quasi-solid-state electrolyte and a LiNi0.6Co0.2Mn0.2O2 cathode was assembled, and its cycling characteristics and thermal stability were investigated. Our results reveal that the quasi-solid-state lithium-ion battery exhibits good capacity retention and superior thermal safety to conventional liquid electrolyte-based cells.

MTASpec software for calculating the vibrational IR and Raman spectra of large molecules at ab initio level

Fragmentation-based (FB) methods have been developed for enabling ab initio calculations on large molecules and clusters. However, a well-benchmarked FB-based utility, for calculating the vibrational spectra is not available. The present article reports a software package MTASpec, based on the FB-molecular tailoring approach (MTA), for computing the single point energy followed by vibrational IR and Raman spectra for spatially extended molecular systems. Accuracy and efficiency of MTASpec are assessed vis-à-vis their full calculation counterparts for some medium- to large-sized molecular systems using HF, DFT and MP2 theory employing large basis sets. The code is fully automated for use on linux platform, with the Gaussian suite of software at the back-end. It is envisaged that the MTASpec package would enable spectral studies of molecular systems containing ∼100 atoms and/or ∼10000 basis functions employing correlated theories with computational economy. Program summaryProgram title: MTASpecCPC Library link to program files:https://doi.org/10.17632/m5b5zhxkfh.1Licensing provisions: GPLv3Programming language: Python, Fortran, C and CshExternal routines/libraries: Gaussian software [1]Nature of problem: Single point energy, harmonic IR and Raman spectra computation for spatially extended molecular systems.Solution method: The Fragment-based method, molecular tailoring approach (MTA) is employed for enabling economic and efficient ab initio computation of the vibrational spectra of large molecules and clusters.

Modeling the deployment of a two-link spoke of a large-sized space reflector taking into account the backlash

This article discusses the process of deployment a two-link spoke of a large-sized transformable space-based reflector. In view of the high costs of carrying out field tests, the construction of correct mathematical models is an urgent task. Currently, the creation of large-sized systems is actively developing. Such systems consist of several interconnected links. When delivered to a given orbit, the large-sized system is folded for placement in the launch vehicle. After entering the orbit, it is deployed to the specified operating state. A mathematical model has been developed for the deployment of the spoke, improved in terms of taking into account the separation of parameters depending on the length and time, which makes it possible to study the arising vibrations of the structure. It is important to take into account the backlash in the connections. Even small gaps in the spoke link connections can lead to a manifold increase in the stabilization time of the system. The developed mathematical model makes it possible to consider various conditions for linking links, change the mass-dimensional parameters and materials of the spoke. The results of modeling are presented, showing the correctness of mathematical models. Conclusions are made about the admissibility of using mathematical models for spokes consisting of a larger number of links.

Determining maximum hosting capacity for PV systems in distribution grids

Power systems face increasing challenges on reliable operations due to the widespread distributed generators (DGs), e.g., rooftop PV systems in distribution grids. Characterizing the hosting capacity (HC) is vital for utilities to assess the total amount of distributed generations that a grid can deploy efficiently before upgrading. Some methods conduct extensive simulations for analyzing HC, which lack theoretical guarantees and can be time-consuming. Therefore, there are also methods employing optimization over all necessary operation constraints. However, the complexity and inherent non-convexity lead to non-optimal solutions. To solve the problems, this paper provides a constructive model for HC determination. Based on geometrical understanding, HC solutions are constructed sequentially according to realistic constraints so that the optimal solution can be reserved even with a non-convex model. For practical adaption, the three-phase unbalance condition is considered, and parallel computation is developed for speed-up. The method is validated by extensive numerical results on IEEE standard systems, such as 8-bus and 123-bus radial distribution grids. In comparison, the proposed approach obtains analytical hosting capacity solutions of larger values and uses less computation time. The results of HC verify that the proposed method reserves optimality and thus has good performance in complex scenarios, including more operational constraints, different PV deployment, and larger system size.

Reinforcement Learning for Digital Quantum Simulation.

Digital quantum simulation on quantum computers provides the potential to simulate the unitary evolution of any many-body Hamiltonian with bounded spectrum by discretizing the time evolution operator through a sequence of elementary quantum gates. A fundamental challenge in this context originates from experimental imperfections, which critically limits the number of attainable gates within a reasonable accuracy and therefore the achievable system sizes and simulation times. In this work, we introduce a reinforcement learning algorithm to systematically build optimized quantum circuits for digital quantum simulation upon imposing a strong constraint on the number of quantum gates. With this we consistently obtain quantum circuits that reproduce physical observables with as little as three entangling gates for long times and large system sizes up to 16 qubits. As concrete examples we apply our formalism to a long-range Ising chain and the lattice Schwinger model. Our method demonstrates that digital quantum simulation on noisy intermediate scale quantum devices can be pushed to much larger scale within the current experimental technology by a suitable engineering of quantum circuits using reinforcement learning.

Worldline algorithm by oracle-guided variational autoregressive network

The variational autoregressive network is extended to the Euclidean path integral representation of quantum partition function. An essential challenge is adapting the sequential process of sample generation by an autoregressive network to a nonlocal constraint due to the periodic boundary condition in path integral. An add-on oracle is devised for this purpose, which accurately identifies and stalls unviable configurations as soon as they occur. The oracle enables rejection-free sampling conforming to the periodic boundary condition. As a demonstration, the oracle-guided autoregressive network is applied to obtain variational solutions of quantum spin chains at finite temperatures with relatively large system sizes and numbers of time slicing, and to efficiently compute thermodynamic quantities.

Machine learning implicit solvation for molecular dynamics.

Accurate modeling of the solvent environment for biological molecules is crucial for computational biology and drug design. A popular approach to achieve long simulation time scales for large system sizes is to incorporate the effect of the solvent in a mean-field fashion with implicit solvent models. However, a challenge with existing implicit solvent models is that they often lack accuracy or certain physical properties compared to explicit solvent models as the many-body effects of the neglected solvent molecules are difficult to model as a mean field. Here, we leverage machine learning (ML) and multi-scale coarse graining (CG) in order to learn implicit solvent models that can approximate the energetic and thermodynamic properties of a given explicit solvent model with arbitrary accuracy, given enough training data. Following the previous ML-CG models CGnet and CGSchnet, we introduce ISSNet, a graph neural network, to model the implicit solvent potential of mean force. ISSNet can learn from explicit solvent simulation data and be readily applied to molecular dynamics simulations. We compare the solute conformational distributions under different solvation treatments for two peptide systems. The results indicate that ISSNet models can outperform widely used generalized Born and surface area models in reproducing the thermodynamics of small protein systems with respect to explicit solvent. The success of this novel method demonstrates the potential benefit of applying machine learning methods in accurate modeling of solvent effects for in silico research and biomedical applications.

NAC-TDDFT: Time-Dependent Density Functional Theory for Nonadiabatic Couplings

First-order nonadiabatic coupling (NAC) matrix elements (fo-NACMEs) are the basic quantities in theoretical descriptions of electronically nonadiabatic processes that are ubiquitous in molecular physics and chemistry. Given the large size of systems of chemical interests, time-dependent density functional theory (TDDFT) is usually the first choice of methods. However, the lack of many-electron wave functions in TDDFT renders the formulation of NAC-TDDFT for fo-NACMEs conceptually difficult. Because of this, various variants of NAC-TDDFT have been proposed in the literature from different standing points, including the Hellmann-Feynman-like expression and auxiliary/pseudo-wave function (AWF)-, equation-of-motion (EOM)-, and time-dependent perturbation theory (TDPT)-based formulations. Based on critical analyses, the following conclusions are made here: (1) The Hellmann-Feynman-like expression, which is rooted in exact wave function theory, is hardly useful due to huge demand on basis sets. (2) Although most popular, the AWF variants of NAC-TDDFT are not theoretically founded and become ambiguous particularly for the fo-NACMEs between two excited states, although they do agree with the EOM and TDPT variants under the Tamm-Dancoff approximation. (3) The TDPT variant of NAC-TDDFT is theoretically most rigorous but suffers from numerical instabilities on the one hand and does not differ to a significant extent from the EOM variant on the other hand. (4) As such, the EOM variant of NAC-TDDFT for the fo-NACMEs between the ground and excited states and between two excited states is solely the right choice in practice. These formal analyses are fully supported by numerical experimentations, taking azulene as a showcase. The proper implementation of the EOM variant of NAC-TDDFT is also highlighted, showing that the fo-NACMEs between the ground and excited states and between two excited states are computationally very much the same as the analytic energy gradients of DFT and TDDFT, respectively. Possible future developments of the EOM variant of NAC-TDDFT are also highlighted. Its extensions to spin-adapted open-shell TDDFT and proper treatment of spin-orbit couplings (which are another source of force for electronically nonadiabatic processes) are particularly warranted in the near future.

Digital Orientation of Health Systems in the Post-COVID-19 "New Normal" in the United States: Cross-sectional Survey.

BackgroundAlmost all health systems have developed some form of customer-facing digital technologies and have worked to align these systems to their existing electronic health records to accommodate the surge in remote and virtual care deliveries during the COVID-19 pandemic. Others have developed analytics-driven decision-making capabilities. However, it is not clear how health systems in the United States are embracing digital technologies and there is a gap in health systems’ abilities to integrate workflows with expanding technologies to spur innovation and futuristic growth. There is a lack of reliable and reported estimates of the current and futuristic digital orientations of health systems. Periodic assessments will provide imperatives to policy formulation and align efforts to yield the transformative power of emerging digital technologies.ObjectiveThe aim of this study was to explore and examine differences in US health systems with respect to digital orientations in the post–COVID-19 “new normal” in 2021. Differences were assessed in four dimensions: (1) analytics-oriented digital technologies (AODT), (2) customer-oriented digital technologies (CODT), (3) growth and innovation–oriented digital technologies (GODT), and (4) futuristic and experimental digital technologies (FEDT). The former two dimensions are foundational to health systems’ digital orientation, whereas the latter two will prepare for future disruptions.MethodsWe surveyed a robust group of health system chief executive officers (CEOs) across the United States from February to March 2021. Among the 625 CEOs, 135 (22%) responded to our survey. We considered the above four broad digital technology orientations, which were ratified with expert consensus. Secondary data were collected from the Agency for Healthcare Research and Quality Hospital Compendium, leading to a matched usable dataset of 124 health systems for analysis. We examined the relationship of adopting the four digital orientations to specific hospital characteristics and earlier reported factors as barriers or facilitators to technology adoption.ResultsHealth systems showed a lower level of CODT (mean 4.70) or GODT (mean 4.54) orientations compared with AODT (mean 5.03), and showed the lowest level of FEDT orientation (mean 4.31). The ordered logistic estimation results provided nuanced insights. Medium-sized (P<.001) health systems, major teaching health systems (P<.001), and systems with high-burden hospitals (P<.001) appear to be doing worse with respect to AODT orientations, raising some concerns. Health systems of medium (P<.001) and large (P=.02) sizes, major teaching health systems (P=.07), those with a high revenue (P=.05), and systems with high-burden hospitals (P<.001) have less CODT orientation. Health systems in the midwest (P=.05) and southern (P=.04) states are more likely to adopt GODT, whereas high-revenue (P=.004) and investor-ownership (P=.01) health systems are deterred from GODT. Health systems of a medium size, and those that are in the midwest (P<.001), south (P<.001), and west (P=.01) are more adept to FEDT, whereas medium (P<.001) and high-revenue (P<.001) health systems, and those with a high discharge rate (P=.04) or high burden (P=.003, P=.005) have subdued FEDT orientations.ConclusionsAlmost all health systems have some current foundational digital technological orientations to glean intelligence or service delivery to customers, with some notable exceptions. Comparatively, fewer health systems have growth or futuristic digital orientations. The transformative power of digital technologies can only be leveraged by adopting futuristic digital technologies. Thus, the disparities across these orientations suggest that a holistic, consistent, and well-articulated direction across the United States remains elusive. Accordingly, we suggest that a policy strategy and financial incentives are necessary to spur a well-visioned and articulated digital orientation for all health systems across the United States. In the absence of such a policy to collectively leverage digital transformations, differences in care across the country will continue to be a concern.

Fast and robust all-electron density functional theory calculations in solids using orthogonalized enriched finite elements

We present a computationally efficient approach to perform systematically convergent real-space all-electron Kohn-Sham DFT calculations for solids using an enriched finite element (FE) basis. The enriched FE basis is constructed by augmenting the classical FE basis with atom-centered numerical basis functions, comprising of atomic solutions to the Kohn-Sham problem. Notably, to improve the conditioning, we orthogonalize the enrichment functions with respect to the classical FE basis, without sacrificing the locality of the resultant basis. In addition to improved conditioning, this orthogonalization procedure also renders the overlap matrix block-diagonal, greatly simplifying its inversion. Subsequently, we use a Chebyshev polynomial based filtering technique to efficiently compute the occupied eigenspace in each self-consistent field iteration. We demonstrate the accuracy and efficiency of the proposed approach on periodic unit-cells and supercells. The benchmark studies show a staggering $130\times$ speedup of the orthogonalized enriched FE basis over the classical FE basis. We also present a comparison of the orthogonalized enriched FE basis with the LAPW+lo basis, both in terms of accuracy and efficiency. Notably, we demonstrate that the orthogonalized enriched FE basis can handle large system sizes of $\sim$10,000 electrons.

Riemann surface crossover for the spectral gaps of open TASEP

We consider the totally asymmetric simple exclusion process (TASEP) with open boundaries, at the edge of the maximal current (MC) phase. Using analytic continuations from the known stationary eigenvalue, we obtain exact expressions for the spectral gaps in the limit of large system size. The underlying Riemann surface, generated by modified Lambert functions, interpolates between the one for periodic TASEP and the one for open TASEP in the MC phase.

An extended recursive decomposition algorithm for dynamic seismic reliability evaluation of lifeline networks with dependent component failures

An extended recursive decomposition algorithm (e-RDA) is proposed to evaluate the dynamic seismic reliabilities of the complex and/or large-sized lifeline network systems with dependent component failures. To effectively analyze statistical dependence of failures of system components, multivariate extreme value response distributions of the components and joint occurrence probabilities of the (disjoint) shortest paths and cuts of systems, a multivariate Gumbel Copula is developed and introduced into the original recursive decomposition algorithm (RDA). Based on the established joint occurrence probability formula of the paths and cuts, two techniques that may accelerate convergence of the upper and lower reliability bounds of complex and/or large-sized systems are developed and embedded into the RDA. Illustrative examples are presented to demonstrate the accuracy, effectiveness and use of the extended RDA for the node weight (only node-type components may fail), edge weight (only line-type components may fail) and general weight (both node- and line-type components may fail) lifeline network systems.

Accurate and Efficient Computation of Optical Absorption Spectra of Molecular Crystals: The Case of the Polymorphs of ROY.

When calculating the optical absorption spectra of molecular crystals from first principles, the influence of the crystalline environment on the excitations is of significant importance. For such systems, however, methods to describe the excitations accurately can be computationally prohibitive due to the relatively large system sizes involved. In this work, we demonstrate a method that allows optical absorption spectra to be computed both efficiently and at high accuracy. Our approach is based on the spectral warping method successfully applied to molecules in solvent. It involves calculating the absorption spectrum of a supercell of the full molecular crystal using semi-local time-dependent density functional theory (TDDFT), before warping the spectrum using a transformation derived from smaller-scale semi-local and hybrid TDDFT calculations on isolated dimers. We demonstrate the power of this method on three polymorphs of the well-known color polymorphic compound ROY and find that it outperforms both small-scale hybrid TDDFT dimer calculations and large-scale semi-local TDDFT supercell calculations, when compared to the experiment.

Stochasticity helps to navigate rough landscapes: comparing gradient-descent-based algorithms in the phase retrieval problem

In this paper we investigate how gradient-based algorithms such as gradient descent (GD), (multi-pass) stochastic GD, its persistent variant, and the Langevin algorithm navigate non-convex loss-landscapes and which of them is able to reach the best generalization error at limited sample complexity. We consider the loss landscape of the high-dimensional phase retrieval problem as a prototypical highly non-convex example. We observe that for phase retrieval the stochastic variants of GD are able to reach perfect generalization for regions of control parameters where the GD algorithm is not. We apply dynamical mean-field theory from statistical physics to characterize analytically the full trajectories of these algorithms in their continuous-time limit, with a warm start, and for large system sizes. We further unveil several intriguing properties of the landscape and the algorithms such as that the GD can obtain better generalization properties from less informed initializations.

One-dimensional long-range Falikov-Kimball model: Thermal phase transition and disorder-free localization

Disorder or interactions can turn metals into insulators. One of the simplest settings to study this physics is given by the Falikov-Kimball model, which describes itinerant fermions interacting with a classical Ising background field. Despite the translational invariance of the model, inhomogenous configurations of the background field give rise to effective disorder physics which lead to a rich phase diagram in two (or more) dimensions with finite temperature charge density wave (CDW) transitions and interaction-tuned Anderson versus Mott localized phases. Here, we propose a generalised Falikov-Kimball model in one dimension with long-range interactions which shows a similarly rich phase diagram. We use an exact Markov Chain Monte Carlo method to map the phase diagram and compute the energy resolved localisation properties of the fermions. We compare the behaviour of this transitionally invariant model to an Anderson model of uncorrelated binary disorder about a background CDW field which confirms that the fermionic sector only fully localizes for very large system sizes.

Pairing and Pair Superfluid Density in One-Dimensional Two-Species Fermionic and Bosonic Hubbard Models.

We use unbiased computational methods to elucidate the onset and properties of pair superfluidity in two-species fermionic and bosonic systems with onsite interspecies attraction loaded in a uniform, i.e., with no confining potential, one-dimensional optical lattice. We compare results from quantum MonteCarlo (QMC) and density matrix renormalization group (DMRG), emphasizing the one-to-one correspondence between the Drude weight tensor, calculated with DMRG, and the various winding numbers extracted from the QMC. Our results show that, for any nonvanishing attractive interaction, pairs form and are the sole contributors to superfluidity; there are no individual contributions due to the separate species. For weak attraction, the pair size diverges exponentially, i.e., Bardeen-Cooper-Schrieffer (BCS) pairing, requiring huge systems to bring out the pair-only nature of the superfluid. This crucial property is largely overlooked in many studies, thereby misinterpreting the origin and nature of the superfluid. We compare and contrast this with the repulsive case and show that the behavior is very different, contradicting previous claims about drag superfluidity and the symmetry of properties for attractive and repulsive interactions. Finally, our results show that the situation is similar for soft-core bosons: superfluidity is due only to pairs, even for the smallest attractive interaction strength compatible with the largest system sizes that we could attain.

Cost-efficient reconfigurable geometrical bus interconnection system for many-core platforms

System-on-chip (SoC) embedded computing platforms can support a wide range of next generation embedded artificial intelligence and other computationally intensive applications. These platforms require cost effective interconnection network. Network-on-chip has been widely used today for on-chip interconnection. However, it is still considered expensive for large system sizes. As full bus-based interconnection has high number of bus connections, reduced bus connections might offer considerable implementation economies with relatively small design cost for field programmable gate arrays (FPGAs) based embedded platforms. In this paper, we propose a cost efficient generalized reconfigurable bus-based interconnection for many-core system with reduced number of bus connections. We generalize the system with b =min {n,m}/k number of interconnect buses in which where n is the number of processor cores, m is the number of memory-modules and k is the general bus reduction factor. We present four geometrical interconnect configurations and provide their characterization in terms of memory bandwidth, cost per bandwidth and bus fault tolerance for various system sizes. Our results show that these configurations provide reduced cost per bandwidth and can achieve higher system throughput with bus cache.

Methodical approach to estimation of energy characteristics of antenna arrays of space communication means of decimeter wavelengths

The article describes a sequence of mathematical and logical operations for determining the energy and mass-dimensional characteristics of an active phased array antenna operating in the decimeter wavelength range of space communications necessary to ensure continuous operation for a given time. When designing large-size antenna systems, a preliminary assessment of the mass-size, energy and thermal parameters is required to assess the feasibility of the project. In this regard, the article uses a mathematical apparatus that includes a model and methods that allow determining the tactical and technical characteristics of satellite communication systems at the design or development stage. The novelty of the developed method consists in assessing the potential capabilities of spacecraft of satellite communication and retransmission systems, taking into account the energy and weight and size characteristics of their antenna systems and in determining the degree of implementation of the main modes of operation of Earth stations at the design or development stage for the specified operating modes of satellite communication systems.

Evaluation of Model Calibration Method for Simulation Performance of a Public Hospital in Brazil

This work presents an extensive study on methodologies to calibrate electric energy consumption in buildings. A comparison between several calibration methodologies shows different approaches addressing the same issue, suggesting a lack of a unique methodology that is reproducible for every building. Additionally, no methodology fits the Brazilian public context, such as the predominance of Unitary Air Conditioning Systems (UACS) and buildings which have operated for more than 30 years. A new calibration methodology for performance simulation is proposed to deal with such features. The methodology is separated into two evidence-based steps according to the size of the Heating, Ventilation and Air Conditioning (HVAC) systems used to control buildings’ indoor environments: the first step is dedicated to calibrating medium- and large-sized HVAC systems, and the second step is dedicated to calibrating small-sized HVAC systems. University Hospital of University of São Paulo (UH-USP) is used as a test bed to implement the proposed methodology. Accuracy indicators show the efficiency of the methodology in terms of calibrating a simulation of the whole UH-USP building and Chilled Water Plant on a monthly basis in terms of accuracy and the time needed to perform the calibration. However, regarding simulation of UACS, the application of the methodology was inconclusive. This study leaves open the question of the trade-off between increasing model outcome accuracy and the strictness of accuracy indicators applied to UACS and poorly automated large-sized air conditioners.