Edge Constraints Research Articles

Gaplessness in Chinese relative clauses

This study posits the existence of a covert gap in Chinese gapless relative clauses, seeking to elucidate how this gap, created by the omission of certain sentential elements, can be recovered via coercion from the relative head noun. This mechanism is elucidated by analyzing the relativization process in prepositional phrases and serial verb constructions. Within prepositional phrases, two subtypes are discerned. The first subtype involves the omission of the preposition, a process termed Preposition Ghosting, which occurs to avoid a violation of the Edge Constraint. Conversely, in the second subtype, the preposition remains in situ without undergoing direct relativization, attributable to its less straightforwardly inferable thematic relationship with the head noun. In serial verb constructions, the process of relativization may lead to the omission of either the first or second verb, with the omitted verb subsequently recoverable through coercion from the relative head noun. The findings suggest that ‘gapless’ relative clauses in Chinese are, in fact, ‘gapped’, which highlight syntactic universals that transcend surface-level forms and meanings.

CMN: a co-designed neural architecture search for efficient computing-in-memory-based mixture-of-experts

Artificial intelligence (AI) has experienced substantial advancements recently, notably with the advent of large-scale language models (LLMs) employing mixture-of-experts (MoE) techniques, exhibiting human-like cognitive skills. As a promising hardware solution for edge MoE implementations, the computing-in-memory (CIM) architecture collocates memory and computing within a single device, significantly reducing the data movement and the associated energy consumption. However, due to diverse edge application scenarios and constraints, determining the optimal network structures for MoE, such as the expert’s location, quantity, and dimension on CIM systems remains elusive. To this end, we introduce a software-hardware co-designed neural architecture search (NAS) framework, CIM-based MoE NAS (CMN), focusing on identifying a high-performing MoE structure under specific hardware constraints. The results of the NYUD-v2 dataset segmentation on the RRAM (SRAM) CIM system reveal that CMN can discover optimized MoE configurations under energy, latency, and performance constraints, achieving 29.67× (43.10×) energy savings, 175.44×(109.89×) speedup, and 12.24× smaller model size compared to the baseline MoE-enabled Visual Transformer, respectively. This co-design opens up an avenue toward high-performance MoE deployments in edge CIM systems.

PathLAD+: Towards effective exact methods for subgraph isomorphism problem

The subgraph isomorphism problem (SIP) is a challenging problem with wide practical applications. In the last decade, despite being a theoretical hard problem, researchers designed various algorithms for solving SIP. In this work, we propose five main strategies and develop an improved exact algorithm for SIP. First, we design a probing search procedure to try whether the search procedure can successfully obtain a solution at first sight. Second, we design a novel matching ordering strategy as a value-ordering heuristic, which uses some useful information obtained from the probing search procedure to preferentially select some promising target vertices. Third, we discuss the characteristics of different propagation methods in the context of SIP and present an adaptive propagation method to make a good balance between these methods. Moreover, to further improve the performance of solving large graphs, we propose an enhanced implementation of the edge constraint method and a domain limitation strategy, which aims to accelerate the search process. Experimental results on a broad range of classic and graph-database benchmarks show that our proposed algorithm performs better than several state-of-the-art algorithms for the SIP.

WLR-Net: An Improved YOLO-V7 With Edge Constraints and Attention Mechanism for Water Leakage Recognition in the Tunnel

WLR-Net: An Improved YOLO-V7 With Edge Constraints and Attention Mechanism for Water Leakage Recognition in the Tunnel

Building Shadow Detection Based on Improved Quick Shift Algorithm in GF‐2 Images

Shadows in remote sensing images contain crucial information about various features on the ground. In this study, a method for detecting building shadows in GF‐2 images based on improved quick shift was proposed. First, six feature variables were constructed: first principal component (PC1), brightness component (I), normalized difference shadow index (NDSI), morphological shadow index (MSI), normalized difference water index (NDWI), and normalized difference vegetation index (NDVI). Then, the image was segmented to obtain homogeneous objects, which were then classified using a random forest model. Two improvements were added to the quick shift algorithm: using PC1, I, and MSI as input data instead of RGB images; and adding Canny edge constraints. Validation in six research areas yields Kappa coefficients of 0.928, 0.896, 0.89, 0.913, 0.879, and 0.909, confirming method feasibility. In addition, comparative experiments demonstrate its effectiveness and robustness across different land cover types while mitigating the segmentation scale effect.

A CNN- and Transformer-Based Dual-Branch Network for Change Detection with Cross-Layer Feature Fusion and Edge Constraints

Change detection aims to identify the difference between dual-temporal images and has garnered considerable attention over the past decade. Recently, deep learning methods have shown robust feature extraction capabilities and have achieved improved detection results; however, they exhibit limitations in preserving clear boundaries for the identified regions, which is attributed to the inadequate contextual information aggregation capabilities of feature extraction, and fail to adequately constrain the delineation of boundaries. To address this issue, a novel dual-branch feature interaction backbone network integrating the CNN and Transformer architectures to extract pixel-level change information was developed. With our method, contextual feature aggregation can be achieved by using a cross-layer feature fusion module, and a dual-branch upsampling module is employed to incorporate both spatial and channel information, enhancing the precision of the identified change areas. In addition, a boundary constraint is incorporated, leveraging an MLP module to consolidate fragmented edge information, which increases the boundary constraints within the change areas and minimizes boundary blurring effectively. Quantitative and qualitative experiments were conducted on three benchmarks, including LEVIR-CD, WHU Building, and the xBD natural disaster dataset. The comprehensive results show the superiority of the proposed method compared with previous approaches.

A study on static bending behavior of partially elastically supported functionally graded plate with porous voids and prediction of deformation through deep learning

This research explores the static bending behaviour of functionally graded rectangular plates with porous voids. It addresses maximum deformation and static bending factors under uniform-pressure, considering variables such as porous-void distributions, full and partial elastic foundations, and edge constraints. The Rayleigh-Ritz method combined with algebraic polynomials is employed to obtain the numerical solutions. The convergence test shows computing efficiency, while the validation tests against public data and ANSYS findings verify the accuracy of the present numerical model. Additionally, this research presents a deep learning-based Artificial-Neural-Network model for deformation prediction to enhance the depth of the analysis without extensive numerical simulations.

Bayesian network structure learning with a new ensemble weights and edge constraints setting mechanism

Bayesian networks (BNs) are highly effective in handling uncertain problems, which can assist in decision-making by reasoning with limited and incomplete information. Learning a faithful directed acyclic graph (DAG) from a large number of complex samples of a joint distribution is currently a challenging combinatorial problem. Due to the growing volume and complexity of data, some Bayesian structure learning algorithms are ineffective and lack the necessary precision to meet the required needs. In this paper, we propose a new PCCL-CC algorithm. To ensure the accuracy of the network structure, we introduce the new ensemble weights and edge constraints setting mechanism. In this mechanism, we employ a method that estimates the interaction between network nodes from multiple perspectives and divides the learning process into multiple stages. We utilize an asymmetric weighted ensemble method and adaptively adjust the network structure. Additionally, we propose a causal discovery method that effectively utilizes the causal relationships among data samples to correct the network structure and mitigate the influence of Markov equivalence classes (MEC). Experimental results on real datasets demonstrate that our approach outperforms state-of-the-art methods.

Illustrating the importance of edge constraints in backbones of bipartite projections.

Bipartite projections (e.g., event co-attendance) are often used to measure unipartite networks of interest (e.g., social interaction). Backbone extraction models can be useful for reducing the noise inherent in bipartite projections. However, these models typically assume that the bipartite edges (e.g., who attended which event) are unconstrained, which may not be true in practice (e.g., a person cannot attend an event held prior to their birth). We illustrate the importance of correctly modeling such edge constraints when extracting backbones, using both synthetic data that varies the number and type of constraints, and empirical data on children's play groups. We find that failing to impose relevant constraints when the data contain constrained edges can result in the extraction of an inaccurate backbone. Therefore, we recommend that when bipartite data contain constrained edges, backbones be extracted using a model such as the Stochastic Degree Sequence Model with Edge Constraints (SDSM-EC).

Effect of constraint relaxation on the minimum vertex cover problem in random graphs.

A statistical-mechanical study of the effect of constraint relaxation on the minimum vertex cover problem in Erdős-Rényi random graphs is presented. Using a penalty-method formulation for constraint relaxation, typical properties of solutions, including infeasible solutions that violate the constraints, are analyzed by means of the replica method and cavity method. The problem involves a competition between reducing the number of vertices to be covered and satisfying the edge constraints. The analysis under the replica-symmetric (RS) ansatz clarifies that the competition leads to degeneracies in the vertex and edge states, which determine the quantitative properties of the system, such as the cover and penalty ratios. A precise analysis of these effects improves the accuracy of RS approximation for the minimum cover ratio in the replica symmetry-breaking (RSB) region. Furthermore, the analysis based on the RS cavity method indicates that the RS/RSB boundary of the ground states with respect to the mean degree of the graphs is expanded, and the critical temperature is lowered by constraint relaxation.

Wave interaction with multiple adjacent floating solar panels with arbitrary constraints

The problem of wave interaction with multiple adjacent floating solar panels with arbitrary types and numbers of constraints is considered. All the solar panels are assumed to be homogeneous, with the same physical properties, as well as modeled by using the Kirchhoff-Love plate theory. The motion of the fluid is described by the linear velocity potential theory. The domain decomposition method is employed to obtain the solutions. In particular, the entire fluid domain is divided into two types, the one below the free surface, and the other below elastic plates. The velocity potential in the free surface domain is expressed into a series of eigenfunctions. By contrast, the boundary integral equation and the Green function are employed to construct the velocity potential of fluid beneath the entire elastic cover, with unknowns distributed along two interfaces and jumps of physical parameters of the plates. All these unknowns are solved from the system of linear equations, which is established from the matching conditions of velocity potentials and edge conditions. This approach is confirmed with much higher computational efficiency compared with the one only involving eigenfunction expansion for the fluid beneath each plate. Extensive results and discussions are provided for the reflection and transmission coefficients of water waves, maximum deflection, and principal strain of the elastic plates; especially, the influence of different types and numbers of edge constraints are investigated in detail.

Concurrent process and feedrate scheduling with convoluted basis functions and its application to fluid jet polishing

Non-traditional laser and fluid jet processes exhibit time-dependent material removal characteristics. The feedrate profile must be planned carefully along the toolpath for accurate surface profile generation while ensuring that the kinematic limits of machine tools are not violated. Conventional methods iteratively solve a deconvolution/convolution problem on the dwell-time density (reciprocal of the feedrate profile) that is computationally heavy, may leave significant residual processing errors, and even generate infeasible feed profiles with the manufacturing equipment. This paper proposes a novel approach that fully addresses the shortcomings above. Dwell-time density is first expressed as a continuous B-spline profile. The associated dwell basis functions (DBF) are convolved with the process influence function (PIF) to generate new process basis functions (PBF). This approach conveniently allows the posing of the problem as a concurrent linear least-squares problem on the control points shared by the DBFs and PBFs while ensuring the numerical stability of the solution and smoothness of the feed profile. To mitigate excessive acceleration peaks and any ringing effect around the edges of the toolpath, this paper also presents methodologies for stabilizing the scheduled feedrate profile by introducing knot vector adjustments (adaptive knot dropping) and linear edge constraints. The effectiveness of the proposed method is demonstrated and validated through simulation case studies and experimentally in fluid jet processing of precision optics. Results indicate that the proposed technique overcomes the limitations of conventional strategies and allows high-frequency surface components beyond the first zero-power frequency of the process footprint to be tracked while still generating a smooth feed profile within the acceleration limits of a machine tool. This ability stems from the localization characteristics associated with the basis functions. By improving the accuracy of high-frequency components, the proposed method exhibits the potential to fabricate topographies with sharper edges, which has been a challenge for conventional techniques.

Bending analysis of functionally graded plates resting on elastic foundation: A Rayleigh-Ritz approach and ANN method

This study presents a comprehensive investigation into the static bending characteristics of rectangular plates constructed from functionally graded materials resting on an elastic foundation. The research examines numerical factors that include central deflection, bending, and normal stresses, in functionally graded rectangular plates (FGRP) subjected to mixed edge constraints under various loading conditions. Classical plate theory is employed, and a computationally efficient mathematical expression for bending analysis is proposed, utilizing the energy principle. The Rayleigh-Ritz method with an algebraic function is applied to solve the governing equation, and the computation is performed using MATLAB. The study investigates the impact of aspect ratios, material property exponents, and the moduli of the Winkler and Pasternak foundations on the numerical factors of embedded FGRPs. The findings reveal that material property exponents and aspect ratios significantly affect numerical factors. The Pasternak foundation moduli demonstrate a more pronounced influence compared to the Winkler foundation moduli. This research provides valuable insights about the static analysis of FGRPs on Winkler-Pasternak elastic foundations, offering practical applications in the design and analysis of structures with similar configurations. Additionally, a salient aspect of this research is in the implementation of an artificial neural network (ANN) for the purpose of predicting the central deformation of FGRP while manipulating the input parameters. This approach is especially advantageous in engineering applications as it facilitates precise predictions for any set of input parameters without using complex computational work.

Numerical Simulation Study on Mechanical Bearing Behavior of Arch Steel–Concrete Composite Sandwich Roof

In order to study the mechanical bearing behavior of arched sandwich roof structures, a full combination and independent action mode of concrete sandwich composite panels was constructed using the finite element method, and an arched steel–concrete composite sandwich roof with a span of 18 m was subjected to a numerical simulation test under a full-span vertical uniformly distributed load, with the bearing characteristics of the arched sandwich roof discussed in depth. The results show that the cross-sections of l/16 and l/2 of the elliptical arch sandwich roof are weak sections, and the tensile cracking of concrete appears for the first time in the upper and lower wythes of the elliptical arch sandwich roof, the von Mises stress level of the lower wythe of the l/16 section is higher under the ultimate load, and the roof shows four-part form failure characteristics. With the expansion of the cracking range of the upper and lower concrete wythes of the steel–concrete composite sandwich arch roof, the load–displacement curve of the roof structure does not decrease significantly, and the bearing capacity of the structure is high and the vertical deformation is small. The steel–concrete composite segment at the end of the roof effectively strengthens the edge constraint of the roof and improves the integrity of the sandwich roof. The upper and lower concrete wythes of the sandwich roof show a fully combined action mode in the elastic working stage and, when the concrete cracks, it shows a partial combined action mode.

From Noise Addition to Denoising: A Self-Variation Capture Network for Point Cloud Optimization.

Point clouds obtained from 3D scanners are often noisy and cannot be directly used for subsequent high-level tasks. In this article, we propose a novel point cloud optimization method capable of denoising and homogenizing point clouds. Our idea is based on the assumption that the noise is generally much smaller than the effective signal. We perform noise perturbation on the noisy point cloud to get a new noisy point cloud, called self-variation point cloud. The noisy point cloud and self-variation point cloud have different noise distribution, but the same point cloud distribution. We compute the potential commonality between two noisy point clouds to obtain a clean point cloud. To implement our idea, we propose a Self-Variation Capture Network (SVCNet). We perturb the point cloud features in the latent space to obtain self-variation feature vectors, and capture the commonality between two noisy feature vectors through the feature aggregation and averaging. In addition, an edge constraint module is introduced to suppress low-pass effects during denoising. Our denoising method does not take into account the noise characteristics, and can filter the drift noise located on the underlying surface, resulting in a uniform distribution of the generated point cloud. The experimental results show that our algorithm outperforms the current state-of-the-art algorithms, especially in generating more uniform point clouds. In addition, extended experiments demonstrate the potential of our algorithm for point clouds upsampling.

The low frequency multi-linear spectrum vibration control study of cylindrical shell through a semi-analytical method

Dynamic vibration absorber is efficient in controlling low frequency vibration of cylindrical shell, which is meaningful for engineering practice. However, the efficiency of designing dynamic vibration absorber using traditional method is low. This paper demonstrates a semi-analytical method for designing dynamic vibration absorber of cylindrical shell. The convergence analysis presented in the research proves the accuracy of the current method when handling different edge constraints through limited truncation numbers. Additionaly, forced vibration and control of multi-line spectrum vibration of cylindrical shell have also been studied. The comparison between FEM and current method shows a significant improvement when handling equivalent mass, which is a key process when designing dynamic vibration absorbers. Finally, an intriguing relationship is discovered when controlling multiple resonance peaks through a single dynamic vibration absorber. The semi-analytical method described in this paper may be beneficial when controlling the vibration of cylindrical shell structures in the further research.

Free Vibration and Dynamical Analyses of FGM Plates with Porosity and Tangential Edge Constraints

Free Vibration and Dynamical Analyses of FGM Plates with Porosity and Tangential Edge Constraints

A Semi-Analytical Method to Design Distributed Dynamic Vibration Absorber of Beam Under General Elastic Edge Constraints

In engineering practice, dynamic vibration absorber is an effective method to control vibration of beam structure. The mathematical model of beam coupled with spring–mass systems under elastic edge constraints is established in this paper. In addition to Fourier cosine series, supplementary sine functions are introduced to represent the displacement function of the beam. The spring–mass system is considered by increasing degrees of freedom when constructing the mass matrix and stiffness matrix. The efficiency of solving equivalent mass is greatly improved in the comparison with FEM, which is important when designing dynamic vibration absorber. Subsequently, we conduct parametric analysis of the control effect about the dynamic vibration absorber based on the current mathematical model. The procedure of designing distributed dynamic vibration absorber of beam structure under various boundary condition is also presented. To verify the validity of the current mathematical model and design procedure, a cantilever beam experiment was conducted. The findings presented in the current research may be useful for researchers in the process of multi-line spectrum vibration control of beam structure.

Performance of Common Scene Stacking Atmospheric Correction on Nonlinear InSAR Deformation Retrieval

Atmospheric Phase Screen (APS) is a major noise that suppresses the accuracy of InSAR deformation time series products. Several correction methods have been developed to perform APS reduction in the InSAR analysis, in which an algorithm called Common Scene Stacking (CSS) method draws wide attention in the community as the method was supposed to effectively separate atmospheric contributions without any external data. CSS was initially proposed for solving linearly interseismic deformation. Whether CSS can be applied in nonlinear deformation cases remains unsolved. In this study, we first conduct a series of data simulations including variable elastic deformation components and also propose an iterative strategy to address the inherent weak edge constraint issues in CSS under different deformation conditions. The results show that signal-to-noise ratio (SNR) is a key parameter affecting the performance of CSS in APS separation. For example, the recovery rate of deformation can generally be greater than 80% from datasets with SNR greater than 10 dB. Our results imply that CSS can favor further improvement of InSAR measurement accuracy. The proposed method in this study was applied to assessing deformation history across the 2020 Mw 5.7 Dingjie earthquake, in which logarithmic postseismic deformation history and coseismic contribution can be successfully retrieved once.

Self-weight and thermal bifurcation analysis of functionally graded vertical annular plates

The present paper thoroughly investigates the influence of structural weight on the thermal buckling of standing circular and annular plates. The structure is composed of symmetrically functionally graded materials (FGM) in terms of its thickness. The first-order theory of shear deformation along with nonlinear strain–displacement equations has been employed to derive the governing equations of the FG plate. First, the asymmetric pre-buckling path of the structure under the influence of weight and heat is calculated by a semi-analytical method. The plate's stability relations are extracted using the adjacent equilibrium criterion and virtual work principle. Then, the stability equations are analyzed by the 2D generalized differential quadrature (GDQ) method. Parametric results are presented based on the analysis of the FGM power-law index, geometric ratios, and edge constraints on the buckling of the vertical plate. The study reveals the critical power-law index and thickness at which the structure becomes unstable, shedding light on the previously neglected calculation of weight-induced buckling in vertical structures. Also, the effect of weight on the buckling mode shapes of heavy plates and disturbing its symmetry is demonstrated. Additionally, this study presents the first investigation into the effect of weight on the thermal stability of annular FG plates.