Explicit Update Research Articles

DR-Label: Label Deconstruction and Reconstruction of GNN Models for Catalysis Systems

Attaining the equilibrium geometry of a catalyst-adsorbate system is key to fundamentally assessing its effective properties, such as adsorption energy. While machine learning methods with advanced representation or supervision strategies have been applied to boost and guide the relaxation processes of catalysis systems, existing methods that produce linearly aggregated geometry predictions are susceptible to edge representations ambiguity, and are therefore vulnerable to graph variations. In this paper, we present a novel graph neural network (GNN) supervision and prediction strategy DR-Label. Our approach mitigates the multiplicity of solutions in edge representation and encourages model predictions that are independent of graph structural variations. DR-Label first Deconstructs finer-grained equilibrium state information to the model by projecting the node-level supervision signal to each edge. Reversely, the model Reconstructs a more robust equilibrium state prediction by converting edge-level predictions to node-level via a sphere-fitting algorithm. When applied to three fundamentally different models, DR-Label consistently enhanced performance. Leveraging the graph structure invariance of the DR-Label strategy, we further propose DRFormer, which applied explicit intermediate positional update and achieves a new state-of-the-art performance on the Open Catalyst 2020 (OC20) dataset and the Cu-based single-atom alloys CO adsorption (SAA) dataset. We expect our work to highlight vital principles for advancing geometric GNN models for catalysis systems and beyond. Our code is available at https://github.com/bowenwang77/DR-Label

Structure-exploiting interior-point solver for high-dimensional entropy-sparsified regression learning

The solution of high-dimensional nonlinear regression problems through standard machine learning approaches often relies on first-order information, due to the numerical and memory challenges arising from the computation of the Hessian matrix and of the higher-order derivatives. While this scenario seems not favorable to second-order methods, here we show that an efficient and modular structure-exploiting interior-point solver can be successfully applied to the recently introduced class of entropy-based methods for regression learning. Specifically, by exploiting the favorable structure of the problem and of the Hessian matrix, we suggest a robust solution strategy based on explicit low-rank updates combined with an iterative Symmetric Quasi-Minimal Residual (SQMR) algorithm to solve the underlying system of linear equations. The results show that the proposed structure-exploiting solver – which relies on the hybrid parallelism and distributed-memory computing paradigm – allows a significant solution time speed-up with respect to a naive solution strategy. Furthermore, through an adequate use of the Message Passing Interface (MPI) and of Open Multi-Processing (OpenMP), the proposed solver enables the solution of large-scale problems on high-performance computing architectures consisting of thousands of compute nodes. The accompanying detailed convergence and performance analyses demonstrate both numerical robustness and high-performance capabilities for increasingly high-dimensional problems.

An Explicit Scheme for Energy-Stable Simulation of Mass-Barrier Collisions with Contact Damping and Dry Friction

Collisions feature in the acoustics of many musical instruments, and as such need to be included in physical models that aim to capture the resulting dynamics. Previous studies have established various ways of constructing energy-stable schemes for numerical modelling of collisions, usually in implicit form. Recent methods use energy quadratisation in the derivation of explicit update forms, which are particularly useful in a real-time synthesis context. This approach is extended here by including contact damping and sliding friction in a 2D one-mass impact oscillator model, which serves as an initial testbed problem. The scheme’s guaranteed passivity is formally shown, and brief simulation and convergence results are included.

A Time-Domain Volume Integral Equation Solver to Analyze Electromagnetic Scattering From Nonlinear Dielectric Objects

A time domain electric field volume integral equation (TD-EFVIE) solver is proposed for analyzing electromagnetic scattering from dielectric objects with Kerr nonlinearity. The nonlinear constitutive relation that relates electric flux and electric field induced in the scatterer is used as an auxiliary equation that complements TD-EFVIE. The ordinary differential equation system that arises from TD-EFVIE’s Schaubert-Wilton-Glisson (SWG)-based discretization is integrated in time using a predictor-corrector method for the unknown expansion coefficients of the electric field. Matrix systems that arise from the SWG-based discretization of the nonlinear constitutive relation and its inverse obtained using the Padé approximant are used to carry out explicit updates of the electric field and the electric flux expansion coefficients at the predictor and the corrector stages of the time integration method. The resulting explicit marching-on-in-time (MOT) scheme does not call for any Newton-like nonlinear solver and only requires solution of sparse and well-conditioned Gram matrix systems at every step. Numerical results show that the proposed explicit MOT-based TD-EFVIE solver is more accurate than the finite-difference time-domain method that is traditionally used for analyzing transient electromagnetic scattering from nonlinear objects.

Convergence analysis of a positivity-preserving numerical scheme for the Cahn-Hilliard-Stokes system with Flory-Huggins energy potential

A finite difference numerical scheme is proposed and analyzed for the Cahn-Hilliard-Stokes system with Flory-Huggins energy functional. A convex splitting is applied to the chemical potential, which in turns leads to the implicit treatment for the singular logarithmic terms and the surface diffusion term, and an explicit update for the expansive concave term. The convective term for the phase variable, as well as the coupled term in the Stokes equation, is approximated in a semi-implicit manner. In the spatial discretization, the marker and cell difference method is applied, which evaluates the velocity components, the pressure and the phase variable at different cell locations. Such an approach ensures the divergence-free feature of the discrete velocity, and this property plays an important role in the analysis. The positivity-preserving property and the unique solvability of the proposed numerical scheme are theoretically justified, utilizing the singular nature of the logarithmic term as the phase variable approaches the singular limit values. An unconditional energy stability analysis is standard, as an outcome of the convex-concave decomposition technique. A convergence analysis with accompanying error estimate is provided for the proposed numerical scheme. In particular, a higher order consistency analysis, accomplished by supplementary functions, is performed to ensure the separation properties of numerical solution. In turn, using the approach of rough and refined error estimates, we are able to derive an optimal rate convergence. To conclude, several numerical experiments are presented to validate the theoretical analysis.

Filtering Methods for Coupled Inverse Problems

In many applications it is required to determine a parameters set which is suitable for competing models. To this end, we are interested in ensemble methods to solve multiobjective optimization problems. Here, the ensemble Kalman Filter method is applied and adapted in order to solve coupled inverse nonlinear problems using a weighted function approach. An analysis of the mean field limit of the ensemble method yields an explicit update formula for the weights. Numerical examples show the improved performance of the proposed method.

The role of storage degradation in energy management problems: An optimal control perspective

The energy management problem of grid-connected storage systems is becoming crucial due to massive integration of renewable energy sources. However, in these problems, the storage degradations are often overlooked while designing the optimal control policy. The key reason behind that is the degradation cost, which occurs at each (dis)charging storage cycle, is hard to evaluate, and it is highly dependent on the grid demand. Thus, integrating them in the energy management problem is a challenge. In this article, we address this challenge, and propose an optimal control update, which solves the energy management problem of a power grid considering the storage degradation cost. We first adopt a multi-storage setup, where each storage system has a constant efficiency function and finite storage capacity (not necessarily the same), and then evaluate necessary optimal conditions based on Pontryagin’s minimum principle. We show that these conditions can further be leveraged to obtain an explicit optimal control update for a family of homogeneous storage systems. Moreover, we derive explicit control updates for the energy management problems, where the storage degradation costs obey either the additive or the multiplicative degradation cost models. Several numerical experiments are performed on the Israel electric grid to demonstrate the covered concepts and the effectiveness of the proposed approach.

Widely Distributed Radar Imaging: Unmediated ADMM Based Approach

This paper presents a novel approach to reconstruct a unique image of an observed scene via synthetic apertures (SA) generated by employing widely distributed radar sensors. The problem is posed as a constrained optimization problem in which the global image which represents the aggregate view of the sensors is a decision variable. While the problem is designed to promote a sparse solution for the global image, it is constrained such that a relationship with local images that can be reconstructed using the measurements at each sensor is respected. Two problem formulations are introduced by stipulating two different establishments of that relationship. The proposed formulations are designed according to consensus ADMM (CADMM) and sharing ADMM (SADMM), and their solutions are provided accordingly as iterative algorithms. We drive the explicit variable updates for each algorithm in addition to the recommended scheme for hybrid parallel implementation on the distributed sensors and a central processing unit. Our algorithms are validated and their performance is evaluated by exploiting the Civilian Vehicles Dome dataset to realize different scenarios of practical relevance. Experimental results show the effectiveness of the proposed algorithms, especially in cases with limited measurements.

Positivity-Preserving Lax–Wendroff Discontinuous Galerkin Schemes for Quadrature-Based Moment-Closure Approximations of Kinetic Models

The quadrature-based method of moments (QMOM) offers a promising class of approximation techniques for reducing kinetic equations to fluid equations that are valid beyond thermodynamic equilibrium. In this work, we study a particular five-moment variant of QMOM known as HyQMOM and establish that this system is moment-invertible over a convex region in solution space. We then develop a high-order discontinuous Galerkin (DG) scheme for solving the resulting fluid system. The scheme is based on a predictor–corrector approach, where the prediction is a localized space-time DG scheme. The nonlinear algebraic system in this prediction is solved using a Picard iteration. The correction is a straightforward explicit update based on the time-integral of the evolution equation, where the space-time prediction replaces all instances of the exact solution. In the absence of limiters, the high-order scheme does not guarantee that solutions remain in the convex set over which HyQMOM is moment-realizable. To overcome this, we introduce novel limiters that rigorously guarantee that the computed solution does not leave the convex set of realizable solutions, thus guaranteeing the hyperbolicity of the system. We develop positivity-preserving limiters in both the prediction and correction steps and an oscillation limiter that damps unphysical oscillations near shocks. We also develop a novel extension of this scheme to include a BGK collision operator; the proposed method is shown to be asymptotic-preserving in the high-collision limit. The HyQMOM and the HyQMOM-BGK solvers are verified on several test cases, demonstrating high-order accuracy on smooth problems and shock-capturing capability on problems with shocks. The asymptotic-preserving property of the HyQMOM-BGK solver is also numerically verified.

Optimal structures for failure resistance under impact

The complex physics and numerous failure modes of structural impact creates challenges when designing for impact resistance. While simple geometries of layered material are conventional, advances in 3D printing and additive manufacturing techniques have now made tailored geometries or integrated multi-material structures achievable. Here, we apply gradient-based topology optimization to the design of such structures. We start by constructing a variational model of an elastic–plastic material enriched with gradient phase-field damage, and present a novel method to efficiently compute its transient dynamic time evolution. We consider a finite element discretization with explicit updates for the displacements. The damage field is solved through an augmented Lagrangian formulation, splitting the operator coupling between the nonlinearity and non-locality. Sensitivities over this trajectory are computed through the adjoint method, and we develop a numerical method to solve the resulting adjoint dynamical system. We demonstrate this formulation by studying the optimal design of 2D solid–void structures undergoing blast loading. Then, we explore the trade-offs between strength and toughness in the design of a spall-resistant structure composed of two materials of differing properties undergoing dynamic impact.

How do implicit and explicit partner evaluations update in daily life? Evidence from the lab and the field.

Evidence suggesting that implicit partner evaluations (IPEs), but not explicit evaluations (EPEs), can predict later changes in satisfaction and relationship status has led researchers to postulate that IPEs must be especially sensitive to relational reward and costs. However, supporting evidence for this assumption remains scarce, and very little is known regarding how IPEs versus EPEs actually update in everyday life. Two studies (one in-lab dyadic interaction study, N = 255, and one 14-day dyadic diary study, N = 348) investigated updating in IPEs and EPEs in the context of real-life relationship experiences. Study 1 revealed that the level of positive and negative experiences that a couple encountered while discussing a divergence of interests in their relationship predicted pre-to-post changes in EPEs, but not in IPEs. Study 2 revealed that IPEs showed less sensitivity to everyday relationship experiences across multiple metrics over the course of 14 days. Specifically, compared with EPEs, IPEs fluctuated less at the within- (vs. between-) person level, showed less-abrupt changes from day-to-day, and had a substantially weaker relationship with same-day positive and negative experiences. Rather than covarying with same-day experiences, IPEs appeared sensitive to relationship experiences aggregated across multiple prior days as well as to highly diagnostic relationship experiences, such as breakup. Consistent with recent advances in social-cognitive research, these findings support a modified account of IPE sensitivity, according to which IPEs show only gradual shifts under everyday circumstances, but more-dramatic shifts under highly diagnostic circumstances. Implications of these findings for close relationships and implicit social cognition research are discussed. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

Knowledge-based strategies for multi-agent teams playing against Nature

We study teams of agents that play against Nature towards achieving a common objective. The agents are assumed to have imperfect information due to partial observability, and have no communication during the play of the game. We propose a natural notion of higher-order knowledge of agents. Based on this notion, we define a class of knowledge-based strategies, and consider the problem of synthesis of strategies of this class. We introduce a multi-agent extension, MKBSC, of the well-known knowledge-based subset construction applied to such games. Its iterative applications turn out to compute higher-order knowledge of the agents. We show how the MKBSC can be used for the design of knowledge-based strategy profiles, and investigate the transfer of existence of such strategies between the original game and in the iterated applications of the MKBSC, under some natural assumptions. We also relate and compare the “intensional” view on knowledge-based strategies based on explicit knowledge representation and update, with the “extensional” view on finite memory strategies based on finite transducers and show that, in a certain sense, these are equivalent.

Resilient and Consistent Multirobot Cooperative Localization With Covariance Intersection

Cooperative localization is fundamental to autonomous multirobot systems, but most algorithms couple inter-robot communication with observation, making these algorithms susceptible to failures in both communication and observation steps. To enhance the resilience of multirobot cooperative localization algorithms in a distributed system, we use covariance intersection to formalize a localization algorithm with an explicit communication update and ensure estimation consistency at the same time. We investigate the covariance boundedness criterion of our algorithm with respect to communication and observation graphs, demonstrating provable localization performance under even sparse communications topologies. We substantiate the resilience of our algorithm as well as the boundedness analysis through experiments on simulated and benchmark physical data against varying communications connectivity and failure metrics. Especially when inter-robot communication is entirely blocked or partially unavailable, we demonstrate that our method is less affected and maintains desired performance compared to existing cooperative localization algorithms.

An Empirical Study of Dependency Downgrades in the npm Ecosystem

In a software ecosystem, a dependency relationship enables a <i>client</i> package to reuse a certain version of a <i>provider</i> package. Packages in a software ecosystem often release versions containing bug fixes, new functionalities, and security enhancements. Hence, updating the provider version is an important maintenance task for client packages. Despite the number of investigations about dependency updates, there is a lack of studies about dependency downgrades in software ecosystems. A downgrade indicates that the adopted version of a provider package is not suitable to the client package at a certain moment. In this paper, we investigate downgrades in the <inline-formula><tex-math notation="LaTeX">${\sf npm}$</tex-math></inline-formula> ecosystem. We address three research questions. In our first RQ, we provide a list of the reasons behind the occurrence of downgrades. Our manual analysis of the artifacts (e.g., release notes and commit messages) of a package code repository identified two categories of downgrades according to their rationale: reactive and preventive. The reasons behind reactive downgrades are defects in a specific version of a provider, unexpected feature changes in a provider, and incompatibilities. In turn, preventive downgrades are an attempt to avoid issues in future releases. In our second RQ, we investigate how the versioning of dependencies is modified when a downgrade occurs. We observe that 49 percent of the downgrades are performed by replacing a range of acceptable versions of a provider by a specific old version. This observation suggests that client packages have the tendency to become more conservative regarding the update of their providers after a downgrade. Also, 48 percent of the downgrades reduce the provider version by a minor level (e.g., from 2.1.0 to 2.0.0). This observation indicates that client packages in <inline-formula><tex-math notation="LaTeX">${\sf npm}$</tex-math></inline-formula> should be cautious when updating minor releases of the provider (e.g., by prioritizing tests). Finally, in our third RQ we observe that 50 percent of the downgrades are performed at a rate that is 2.6 times as slow as the median time-between-releases of their associated client packages. We also observe that downgrades that follow an explicit update of a provider package occur faster than downgrades that follow an implicit update. Explicit updates occur when the provider is updated by means of an explicit change to the versioning specification (i.e., the string used by client packages to define the provider version that they are willing to adopt). We conjecture that, due to the controlled nature of explicit updates, it is easier for client packages to identify the provider that is associated with the problem that motivated the downgrade.

An alternating-direction hybrid implicit-explicit finite-difference time-domain method for the Schrödinger equation

This paper proposes a novel hybrid FDTD method for solving the time-dependent Schrödinger equation, which is fundamental for modeling materials and designing nanoscale devices. The wave function is propagated on nonuniform grids by applying explicit updates in part of the grid and implicit updates elsewhere. The latter are based on the Alternating-Direction Implicit (ADI) scheme while the former are constructed with a central difference for the time derivative. A rigorous stability analysis proves that spatial steps can be selectively removed from the stability criterion thus combining the unconditional stability of the ADI scheme with fast explicit calculations. The scheme excels in its flexibility by efficiently discretizing and balancing explicit with implicit updates, as such expediting the computations. Moreover, it retains the linear complexity of explicit schemes with respect to space and time, making it especially scalable to numerically large problems. Several numerical experiments, including a laterally tunnel-coupled quantum wire and a nanowire double-barrier resonant-tunneling diode, show the validity of the scheme by demonstrating its high accuracy and decreased CPU time compared to traditional methods.

Comparison of implicit and explicit numerical integration schemes for a bounding surface soil model without elastic range

Standard integration schemes for rate constitutive equations were designed within the classical theory of plasticity. Consequently, they rely on the assumption that a yield criterion defines a range of purely elastic material behaviour. Many constitutive models for non-cohesive soils discard that assumption. They account for inelastic deformations without employing a yield criterion. This simplifies the formulation of the models but raises questions concerning their numerical implementation. In particular, it is not fully clear which algorithmic method is most appropriate for the numerical integration of constitutive models without elastic range. To investigate this, two different stress point algorithms for a critical state bounding surface model for sands were developed and implemented. The explicit method employs substepping to automatically control the local error. The implicit method updates stresses and state variables through a local Newton iteration, the Jacobian of which is computed by numerical differentiation. The two algorithms were compared by means of calculations at integration point level as well as with respect to a boundary value problem. The results show that for a given level of accuracy, the explicit update procedure is significantly more efficient than the implicit one. This holds regardless of initial state parameters and input strain increment magnitudes.

Zero-order moving horizon estimation for large-scale nonlinear processes

Moving Horizon Estimation (MHE) is an optimization-based approach to nonlinear state estimation where the state estimate is obtained as the solution of a nonlinear optimization problem. Especially for large-scale nonlinear systems, the computational burden associated with the numerical solution of nonlinear optimization problems poses a major challenge when applying MHE in practice. To alleviate the computational effort, we propose an inexact, but computationally less expensive variant of the Gauss–Newton algorithm tailored to the nonlinear least-squares problem arising from the MHE formulation. The proposed algorithm combines two ideas: On the one hand, it uses the fact that the arrival cost matrix appears naturally within the Gauss–Newton Hessian approximation in order to avoid any explicit arrival cost update. On the other hand, the method follows a zero-order optimization approach, where fixed sensitivity approximations are used in order to reduce the number of sensitivity evaluations, while accepting some loss of optimality. The combination of these two ideas allows one to reuse the factorizations of the Hessian blocks associated with each stage, when moving from one MHE problem to the next, therefore significantly reducing the computational complexity. Additionally, a tailored integration method for large-scale stiff systems is proposed that can efficiently propagate approximate forward sensitivities, which are used within the inexact MHE algorithm. Both estimation accuracy and computational efficiency of the proposed method are evaluated by applying it to two case studies: a small-scale batch reactor model and the large-scale dynamic process of acrylic acid production.

Sequential selection for accelerated life testing via approximate Bayesian inference

AbstractAccelerated life testing (ALT) is typically used to assess the reliability of material's lifetime under desired stress levels. Recent advances in material engineering have made a variety of material settings readily available. A critical question is how to efficiently conduct ALT to optimize reliability performance over different material settings. We propose a sequential selection approach to solve this problem. The proposed approach contains (1) a model updating mechanism to incorporate new experimental data in each step, and (2) a design criterion to guide new experiments that maximizes the potential to find the optimal material setting. To guarantee a tractable statistical mechanism for information collection, we develop explicit model parameter update formulas via approximate Bayesian inference. Theories show that our explicit update formulas give consistent parameter estimates. To guarantee that the design criterion in each step can make improvement on the identification of optimal material setting, this paper adopts an expected improvement‐based design criterion for optimizing the material setting under target stress factor levels. We also give a heuristic on this design criterion to justify the statistical consistency of approximate Bayesian estimates. Simulation studies and a case study show that the proposed sequential selection approach can significantly improve the probability of identifying the material alternative with best reliability performance compared to other design approaches.

Default Negation as Explicit Negation plus Update

We argue that under the stable model semantics default negation can be read as explicit negation with update. We show that dynamic logic programming which is based on default negation, even in the heads, can be interpreted in a variant of updates with explicit negation only. As corollaries, we get an easy description of default negation in generalized and normal logic programming where initially negated literals are updated. These results are discussed with respect to the understanding of negation in logic.

A semantic similarity computation method for virtual resources in cloud manufacturing environment based on information content

To ensure the effective and reasonable classification of virtual resources and provide support for the search and matching of required resources that accomplish tasks in cloud manufacturing (CMfg) environments, the semantic similarity computation of virtual resources needs to be promoted for that it is difficult to ensure that a large number of effective features participate in the calculation of semantic similarity and it is difficult to practically apply to efficient classification. However, to promote semantic similarity calculations still have problems such as difficulty in unifying the virtual resource pool and lagging in updating semantic depth calculations. In this paper, we propose a creative approach that can compute the semantic similarity of virtual resources described by video semantic description (VSD) based on information content (IC). In this approach, first, a multi-level progression semantics description framework for virtual resources is proposed with the help of VSD that can provide a new solution because of its explicit description and dynamic update to obtain real and reliable data; second, an event cloud (EC) that serves as the computing scope for semantics similarity calculation of virtual resources is constructed to settle the problem of substantial data growth caused by real-time resource updates on top of the resource manufacturing processes; third, based on an effective solution of corpus dependency and data sparseness, an extended semantic computing model is developed to calculate quantificationally the semantic similarities of virtual resources based on semantic relationships. The results show that compared with classifying virtual resources qualitatively, the semantic similarity computing method that starts from the manufacturing result and considers the comprehensive relationship is more reliable and effective.